Ali Hemani uses marijuana a few times a week. That fact alone, the government says, means he is automatically banned from possessing a firearm under federal law. And because Mr. Hemani admits he owns a gun despite this ban, the government now seeks to prosecute him, imprison him for up to 15 years, and disarm him for life. This case poses the question whether the government’s prosecution of Mr.
Hemani is consistent with the Second Amendment.
I
The federal Gun Control Act prohibits various categories of people from possessing guns. These categories, listed in 18 U. S. C. §922(g), include individuals convicted of crimes “punishable by imprisonment for a term exceeding one year,” §922(g)(1), fugitives from justice, §922(g)(2), and those convicted of domestic violence crimes, §922(g)(9).
This case concerns another of the statute’s provisions— §922(g)(3). Under its terms, anyone who is an “unlawful user of ” or “addicted to” a “controlled substance” is automatically banned from possessing a gun. And for defying this ban, an individual can face up to 15 years in federal prison, §924(a)(8), and disarmament for life, §922(g)(1).1 To define the term “controlled substance,” §922(g)(3) re- lies on the Controlled Substances Act (CSA). Enacted to protect “the health and general welfare of the American people,” 21 U. S. C. §801(2), the CSA sets forth five sched- ules of “controlled substances” subject to varying degrees of regulation, §802. They range from Schedule I drugs with a high potential for abuse and no currently accepted medical uses (like heroin) to Schedule V drugs with the lowest po- tential for abuse and many accepted medical uses (like Robitussin with codeine). §§802, 812. By incorporating the CSA definition of the term “controlled substance,” 18 U. S. C. §922(g)(3) makes it illegal for anyone who unlaw- fully uses any drug found on any of the CSA’s schedules to possess a firearm for any reason, upon threat of imprison- ment and permanent disarmament.
Convictions for violations of §922(g)(3) account for only about 5% of all §922(g) convictions. Congressional Re- search Service, Guns and Drugs: A Brief History of 18 U. S. C. §922(g)(3) (2026). But, the government insists, a conviction is warranted in this case, and the facts giving rise to it unfolded this way.
Mr. Hemani is a dual citizen of the United States and Pa- kistan who was born in Texas and has spent most of his life there. In recent years, he has lived in the Dallas area with his parents and worked a stable job. But, suspecting Mr.
Hemani and his family members of terrorism-related activ- ities, the government conducted a search of the family home in 2022. Throughout the process, Mr. Hemani proved coop- erative. He surrendered a gun he kept in the house and —————— 1 Because Congress adopted §922(g) pursuant to the Commerce Clause, the statute requires the government to show that a firearm possessed in violation of the statute’s terms “has been shipped or transported in in- terstate or foreign commerce.” As that element is not in dispute here, we do not discuss it further.
pointed agents to some marijuana on the property. He also consented to an interview, telling law enforcement agents that he used marijuana “about every other day.” Record on Appeal 381. After the agents found cocaine in his parents’ closet, Mr. Hemani claimed ownership of that as well, though he maintained that his mother had hidden it from him and that he had not used any recently.
More than six months after the search, the government brought a single-charge indictment against Mr. Hemani.
The charge had nothing to do with terrorism—the reason for the search in the first place. Nor did the charge involve possession of cocaine, drug trafficking, or anything like that. Instead, relying solely on his admitted use of marijuana about every other day, the government prosecuted Mr.
Hemani for knowingly possessing a gun in his home while being an “unlawful user” of a controlled substance. Id., at 12. For that alone, the government claimed, Mr. Hemani faced up to 15 years in prison and disarmament for life. No matter that the government did not assert Mr. Hemani was a drug addict.2 No matter that it did not contend his drug use had ever led him to pose a danger to himself or others. No matter, too, that the government did not claim Mr. Hemani had done anything with his gun other than possess it in his home.
Mr. Hemani moved to dismiss the indictment, arguing that the government’s effort to enforce §922(g)(3) against him violated the Second Amendment. The district court granted the motion and, after an unsuccessful appeal to the Fifth Circuit, the government asked us to review the case. We agreed to do so. 607 U. S. 992 (2025).
II
The Second Amendment protects the right of “all Americans” to keep and bear firearms for self-defense. District of —————— 2For this reason, §922(g)(3)’s disarmament of drug “addict[s]” is not before us and we have no occasion to pass on it.
Columbia v. Heller, 554 U. S. 570, 581 (2008). Of course, like most individual rights, the Second Amendment has its limits. Id., at 626. American legislatures have long regulated the possession and use of firearms to some degree.
See New York State Rifle & Pistol Assn., Inc. v. Bruen, 597 U. S. 1, 17 (2022). But when the government crosses the line from permissible regulation into unconstitutional infringement, courts have a duty to say so in the cases before them—no less in the Second Amendment context than in any other. See ibid.; Heller, 554 U. S., at 606.
To determine when the government infringes the Second Amendment, we begin by asking whether the Amendment’s terms cover the conduct in question. Bruen, 597 U. S., at 24. If so, the Constitution “presumptively” protects it. Ibid. To overcome that presumption, the government then bears the burden of showing its regulatory efforts are “consistent with the Nation’s historical tradition of firearm regulation.” Ibid. Our cases demand this attention to history, we have said, because the Second Amendment was designed to codify a “pre-existing” individual right and guard against its later erosion by majoritarian legislation or judicial fiat. Id., at 25 (emphasis deleted). At the same time, we have recognized that “[t]he regulatory challenges posed by firearms today are not always the same” as those earlier generations faced. Id., at 27. Accordingly, to show that a contemporary regulation is consistent with this Nation’s historical tradition of firearm regulation, we do not require the government to point to a “historical twin” or “precis[e] . . . historical precursors.” United States v. Rahimi, 602 U. S. 680, 692 (2024) (internal quotation marks omitted). Instead, we have said, “the appropriate analysis involves considering whether the challenged regulation is consistent with the principles that underpin our regulatory tradition.” Ibid. And, to that end, the government may “reaso[n] by analogy,” showing that its contemporary regulation is “relevantly similar” to ones “well-established” in the Nation’s history. Bruen, 597 U. S., at 28–30.
While we have not yet had cause to “exhaustive[ly] survey” the features that may render a modern law “relevantly similar” to historical ones, we have said two play a “‘central’” role. Id., at 29. Call them the “why” and “how.” Ibid.; see also Rahimi, 602 U. S., at 692. The more closely a contemporary law mirrors a well-established historical analogue in purpose and operation, the more likely it is to be upheld. Conversely, the more a modern law diverges from traditional laws in purpose and operation, the less likely it is to survive review. See Bruen, 597 U. S., at 29; Rahimi, 602 U. S., at 692.
The government accepts all this. It agrees, too, that §922(g)(3)’s unlawful user provision burdens conduct presumptively protected by the Second Amendment. After all, that statute bans a class of people including Mr. Hemani from possessing essentially any firearm for any purpose. As a result, the government acknowledges, it has a burden to carry.
The burden the government sets for itself in this case is a considerable one. As the government construes §922(g)(3) and seeks to apply it here, the law automatically bans an individual from possessing a gun from the moment he becomes an unlawful user of any controlled substance until he ceases being one. The law, says the government, doesn’t require anything more. It doesn’t matter what controlled substance an individual uses, in what amounts he does so, or whether his drug use has ever made him a danger to himself or others. It doesn’t even matter why he keeps a gun or how safely he does so. And for violating this automatic ban, the government insists, an individual like Mr. Hemani may be sent to prison for up to 15 years and disarmed for life. To meet its burden of showing a law like that is consistent with the Nation’s tradition of firearm regulation, the government relies on an analogy to what it calls “habitual drunkard” laws. These laws, the government submits, enjoy deep roots in the country’s history and are “relevantly similar” to the regulation it wishes to enforce against Mr. Hemani. Bruen, 597 U. S., at 29.
In truth, the habitual drunkard laws the government invokes fall into three general categories. First are vagrancy laws. Both at the founding and in the decades following it, vagrants—a group that sometimes included habitual drunkards—could be “confine[d] in a workhouse” or “jail[ed].” Brief for United States 19. Second are civil-commitment statutes. Around the same time, many States allowed courts to appoint guardians for various individuals, including habitual drunkards, or authorized their “commit[ment] to asylums.” Id., at 21. Third are surety laws. Under them, judicial officers, again in the founding era and later, could compel habitual drunkards and others to post surety bonds to ensure their good behavior. “A person who failed to post bond,” the government explains, “would be jailed, while a person who posted bond and then misbehaved would forfeit the bond.” Id., at 22.
These laws, the government insists, mirror §922(g)(3)’s unlawful user provision in three important respects. Two might be said to concern the “why” associated with these laws, the last their “how.” First, the government says, historical laws targeted habitual drunkards for the same reason §922(g)(3) targets unlawful users—because they regularly use intoxicants. See id., at 17 (calling habitual drunkards and unlawful users “closely analogous”). Second, the government submits, habitual drunkard laws restricted the liberties of people to the same end §922(g)(3) does—to protect the public from “unusually dangerous” individuals who commit “violent crime[s].” Id., at 11, 13–14, 22, 34. Finally, the government argues, the historical laws it cites operated in practice much like §922(g)(3) works— the former allowed governments to detain people in places where they could not bear arms while §922(g)(3) temporarily disarms unlawful users. All told, the government contends, because historical habitual drunkard laws resemble §922(g)(3)’s unlawful user provision in so many pertinent respects, that provision must be constitutional as applied in this case and others.
We disagree. We appreciate that drugs and guns can sometimes make for a dangerous mix. We appreciate, too, that the government’s effort to analogize a modern statute addressing drug use to historical laws must be approached with a sensitivity to the fact that many drugs well known today were unknown in early America. As we have put it, the Second Amendment “can, and must, apply to circumstances beyond those the Founders specifically anticipated.” Bruen, 597 U. S., at 28. But, even taking all that into account, the government cannot carry the burden it has set for itself. We decide cases “based on the historical record” and arguments “compiled by the parties” before us. Id., at 26, n. 6. And the habitual drunkard laws on which the government relies here differ dramatically from §922(g)(3)’s unlawful user provision on every single metric the government invites us to consider: They targeted different kinds of people, did so for different purposes, and operated in different ways. Whether any one of these problems taken in isolation would prove fatal to the government’s cause, we need not decide. Taken cumulatively, we hold, they certainly do. And, apart from pointing to habitual drunkard laws, the government has not even attempted to prove that any other specific historical principle might justify its prosecution in this case.3 —————— 3As in Bruen, we need not decide today “whether courts should primarily rely on the prevailing understanding of an individual right when the Fourteenth Amendment was ratified in 1868” or when the Bill of Rights was ratified in 1791. New York State Rifle & Pistol Assn., Inc. v. Bruen, 597 U. S. 1, 37–38 (2022). For reasons explored below, the government cannot establish that the Second Amendment at either point in time permitted anything like its prosecution here.
A
Start with the government’s first point of comparison. In its view, the reason why vagrancy, civil-commitment, and surety laws targeted habitual drunkards is “closely analogous” to the reason why §922(g)(3)’s unlawful user provision targets unlawful users: Both groups regularly use intoxicants. Brief for United States 17. It’s a claim difficult to square with the historical record.
Around the time of the founding and for decades following it, a habitual drunkard was, as one court put it, someone who “for any considerable part of his time [was] intoxicated to such a degree as to deprive him of his ordinary reasoning faculties.” In re Tracy, 1 Paige Ch. 580, 582–583 (N. Y. Ch. 1829). A regular or even frequent drinker did not usually fit the bill. A “man who [was] intoxicated or drunk one-half his time” was more like it. See Ludwick v. Commonwealth, 18 Pa. 172, 175 (1851). In an early American medical treatise, Dr. Benjamin Rush, a signer of the Declaration of Independence, offered this description from a habitual drunkard about the depths of his condition: “‘Were a keg of rum in one corner of a room, and were a cannon constantly discharging balls between me and it, I could not refrain from passing before that cannon, in order to get at the rum.’” Medical Inquiries and Observations, Upon the Diseases of the Mind 266 (1812).
Just consider how some of the statutes the government relies on defined the term. To qualify as a habitual drunkard under an early Arkansas law, someone had to drink to such excess that he was “incapable of conducting [his] own affairs.” Ark. Rev. Stat., ch. 78, §1 (1838) (W. Ball & S. Roane eds.) (Ball & Roane). A statute in the Minnesota Territory required an individual to be “mentally incompetent.” Minn. Terr. Rev. Stat., ch. 67, §12 (1851). A Connecticut law spoke of a person who had “lost the power of self-control.” Act of July 25, 1874, §1, in 1874 Conn. Pub. Acts 256. And in a statute governing the District of Columbia, Congress defined a habitual drunkard as “any person who, by the use of intoxicating liquors . . . has lost self-control, or become incapable of proper attention to the care and management of his affairs, or habitually or periodically neglectful thereof, or dangerous to himself or others.” Act of Mar. 30, 1876, §9, 19 Stat. 10–11.
Had habitual drunkard laws applied to those who simply drank regularly, many notable early Americans could have faced trouble. John Adams took “a tankard of hard cider” with his “daily breakfast.” W. Rorabaugh, The Alcoholic Republic 6 (1979) (Rorabaugh). Some say James Madison “consumed a pint of whiskey daily.” D. Okrent, Last Call: The Rise and Fall of Prohibition 8 (2010); but see M. Will- Weber, Mint Juleps with Teddy Roosevelt: The Complete History of Presidential Drinking 29 (2014) (arguing Madison “championed wine . . . as a healthier and more respectable choice”). George Washington often drank three glasses of madeira in the evening—“not enough to be considered a heavy drinker in his day.” Id., at 5. Thomas Jefferson enjoyed “3 or 4 glasses [of wine] at dinner.” J. Gabler, Passions: The Wines and Travels of Thomas Jefferson 223 (1995). In fact, just a few days before the framers signed the Constitution, a farewell party gathered for General Washington at Philadelphia’s City Tavern where 55 guests are said to have ordered 54 bottles of madeira, 60 bottles of wine, 8 bottles of “Old stock,” 22 bottles of porter, 8 bottles of cider, 12 bottles of beer, and 7 large bowls of punch. National Park Service, C. Hershey, Historic Furnishings Plan: City Tavern 64–65 (1974).
There was, in short, a “culture of copious drinking” in early America. D. Korostyshevsky, Incapable of Managing His Estate: Habitual Drunkards and the Expansion of Guardianship in the Nineteenth-Century United States, 43 Law & Hist. Rev. 795, 800 (2025). Indeed, in 1829 the “secretary of war estimated that three-quarters of the nation’s laborers drank . . . at least 4 ounces of distilled spirits” every day. Rorabaugh 15. Even the American Temperance Society called those who drank 12 ounces of hard liquor daily mere “‘occasional drunkards.’” Id., at 11. As the Society saw things, it took 24 ounces to qualify as a “‘confirmed drunkar[d].’” Ibid. Given all this, it seems the government’s historical laws targeted habitual drunkards not merely because they regularly used intoxicants, or even sometimes used them to excess. Instead, those laws focused on habitual drunkards because their drinking rendered them practically incapacitated and incapable of managing their affairs. And that hardly compares to whom §922(g)(3)’s unlawful user provision targets on the government’s account. To be sure, the government construes the term “unlawful user” to embrace only “habitual or regular” users, not those who use drugs irregularly. Brief for United States 23–25. But even on that understanding, §922(g)(3) automatically disarms anyone who regularly uses any amount of any controlled substance for anything other than its “prescribed purpose.” Tr. of Oral Arg. 57. The law, the government insists, does not require it to show that a particular individual is regularly incapacitated, much less incapable of conducting his affairs or a threat to himself or others. Put simply, on the government’s telling, §922(g)(3) sweeps in large numbers of people without regard to whether their substance use has the kind of incapacitating effect on them that historical habitual drunkard laws normally required.
This case illustrates the disconnect. The government considers Mr. Hemani an unlawful user of a controlled substance because he admits to using marijuana about every other day. But how much marijuana does Mr. Hemani use, in what potency, and to what effect? Is he routinely unable to manage his affairs, a risk to himself or his family? Or does he use a mild gummy as a sleep aid a few times a week? We do not know and, the government says, it doesn’t matter. The government asks us to analogize him to a habitual drunkard all the same.
Nor does the government’s theory stop at Mr. Hemani. It extends equally to a husband who regularly takes his wife’s prescription Ambien to sleep and a college student who routinely uses a friend’s Adderall to cram for exams. Id., at 56–58. The drug involved makes no difference. Nor, again, does it matter how much an individual uses or the effects it has on him. That someone regularly uses any substance found on any of the CSA’s five schedules for anything other than its “prescribed purpose” is enough. Id., at 57. Without more, the government asks us to analogize all such persons to habitual drunkards. To state the analogy is to expose its deficiency.
B
This divergence is not the government’s only problem.
The government faces more trouble yet when it comes to its next argument. It contends that §922(g)(3) disarms unlawful drug users to protect the public from “unusually dangerous” individuals who will “misuse . . . firearms” to commit “violent crime[s].” Brief for United States 11, 13–14, 22, 34. And, the government submits, early American vagrancy, civil-commitment, and surety laws shared a similar purpose. But even spotting the government its assertion about §922(g)(3)’s aims (something we will return to later), the government misapprehends the purposes animating the historical analogues it invokes. Usually, they had little to do with protecting the public from categorically violent and unusually dangerous persons.
Begin with the vagrancy laws. As a rule, they targeted individuals who “did not meet the societal expectation of work.” W. Quigley, Reluctant Charity: Poor Laws in the Original Thirteen States, 31 U. Rich. L. Rev. 111, 169 (1997) (Quigley). Routinely, that category included not just habitual drunkards but also “Vagabonds, Common Beggars,” “pipers, fidlers, . . . stubborn servants or children, [and] common nightwalkers.” Act of Oct. 1727, 7 Public Records of the Colony of Connecticut from May, 1726, to May, 1735, Inclusive 127–128 (C. Hoadly ed. 1873) (Hoadly).4 Add to that list perhaps any others who would not “provide for themselves or the support of their families.” Id., at 128. Some laws confined vagrants to workhouses where they could be “se[t] to work.” Ibid.; see also Quigley 156–157, 168–169. Other statutes authorized short jail terms. See, e.g., Hittell 1288. Either way, laws like these might have sought to promote productivity and suppress any number of real or perceived vices. But they scarcely focused on protecting the public from a “categor[y]” of “unusually dangerous” persons. Brief for United States 10–11. A similar story unfolds when it comes to what the government describes as civil-commitment laws. Consider the first such law the government cites: an 1827 Act in the Michigan Territory. Act of Apr. 12, 1827, in 1827 Mich.
Terr. Laws 60. Worried that habitual drunkards “oftentimes . . . waste[d] their estates by excessive drinking . . . and thereby involve[d] themselves and [their] families in distress, misery, and ruin,” that law permitted a probate judge to assign a guardian to a habitual drunkard to “take care of such person, and [his] estate.” Id., at 58, 60. Or consider a law the government cites from later in the 19th century. It authorized courts to confine habitual drunkards —————— 4See also, e.g., Act of June 29, 1700, §2, in 1 Acts and Resolves of the Province of the Mass. Bay 378 (1869) (including “stubborn children or servants”); Act of May 14, 1718, in 2 Laws of N. H. 266 (A. Batchellor ed. 1913) (including “such as neglect their Callings, Mispend what they earn, and do not provide for themselves”); Act of June 10, 1799, §§1, 3, in Laws of the State of N. J. 473–474 (1821) (including those who did “not give a good account of themselves,” who “leave their families to be maintained by the city,” or who “pretend to use . . . crafty science”); 2 Codes and Statutes of the State of Cal. 1288 (T. Hittell ed. 1876) (Hittell) (including “idle or dissolute person[s] . . . who lodg[e] in any barn, shed, shop, outhouse, [or] vessel”).
to asylums in order to “preven[t them] from using intoxicating liquors” and “refor[m]” them. Act of May 1, 1890, §2, in 1890 Iowa Acts 67. By their own terms, laws like these did not seek to protect the public from violence so much as to protect habitual drunkards from themselves and their families from financial devastation.5 Last, turn to the surety laws. In Rahimi, we rejected a facial challenge to §922(g)(8), a provision that prohibits firearm possession by certain individuals subject to domestic violence restraining orders. In reaching that decision, we analogized the modern statute in part to historical suretyof-the-peace laws. 602 U. S., at 695–696. Those laws generally required an individual shown to pose a specific threat of violence to post a bond and pledge to “keep the peace.” Id., at 695–697 (internal quotation marks omitted); see also J. Parker, Conductor Generalis 397 (1788) (Parker) (sureties of the peace applied to those who posed “some present or future danger”). Failure to post a bond meant jail, and disturbing the peace after posting a bond meant forfeiting it. Rahimi, 602 U. S., at 695. Some surety-of-the-peace laws, we held, were designed to prevent “violence” and —————— 5See also, e.g., Ball & Roane 456 (permitting probate courts to appoint guardians for habitual drunkards and others “incapable of conducting their own affairs . . . to provide for the safe keeping of such persons, the maintenance of themselves and their families, and the education of their children”); Minn. Terr. Rev. Stat., ch. 67, §12, p. 278 (allowing a probate judge to appoint guardians for those “mentally incompetent to have the care and management of their own property”); Ga. Code §1803 (1861) (similar); Act of Feb. 21, 1872, §1, in 1872 Ill. Laws 477 (similar); Act of Apr. 1, 1870, §1, in 1869–1870 Cal. Stat. 585 (providing that “[t]he Home for the care of the Inebriate of San Francisco shall always be kept open for the reception and care of inebriates”); Act of Apr. 17, 1873, §3, in 1873 Miss. Laws 62 (explaining that the object of civil confinement was “reformation and recovery from such pernicious habit of drunkenness”); Act of Mar. 5, 1860, §7, in 1860 Md. Laws ch. 386 (permitting confinement of a habitual drunkard as “necessary for his or her complete reformation”); 19 Stat. 10–11 (similar).
“targeted the misuse of firearms,” much as §922(g)(8) seeks to do. Id., at 695–696.
Notably, the government doesn’t rely on surety-of-thepeace statutes in this case. And that makes sense. To warrant the imposition of a surety of the peace, just being a habitual drunkard wasn’t usually enough. So in this case the government directs us to a different kind of surety stat- ute—one that required certain individuals to post sureties of “good behavior.” Brief for United States 22 (internal quotation marks omitted). Under those laws, a judicial officer could impose a surety of good behavior on individuals who threatened a “scandal.” Parker 410. And a scandal could include anything from “haunting bawdy houses” to “evesdropp[ing]” to, yes, being a “common drunkar[d].” Ibid. In fact, one surety statute targeted those who had “a common practice of getting drunk, and prophane cursing and swearing, and blasphemy, to the great dishonour of Almighty God.” 1 Del. Laws 173 (1797). But what does any of this prove? To impose a surety of good behavior did not normally require a showing that an individual posed a threat of violence. Instead, these laws usually sought to protect the community from scandals “against good morals.” 5 N.
Dane, A General Abridgment and Digest of American Law 301 (1824) (emphasis deleted); see also Parker 410. And that is nothing like the purpose the government ascribes to §922(g)(3)’s unlawful user provision.
C
The government’s difficulties do not end with the “why” comparisons it invites us to make. They also extend to the “how.” “Even when a law regulates arms-bearing for a permissible reason,” we have said, “it may not be compatible with the [Second Amendment] if it does so to an extent beyond what was done at the founding.” Rahimi, 602 U. S., at 692. And the way habitual drunkard statutes worked in the past differs significantly from how §922(g)(3)’s unlawful user provision works today.
The historical laws the government identifies usually provided some form of process before an individual lost any of his liberties, even temporarily. Normally, a vagrant could be sent to a workhouse or jail only upon a “conviction.” E.g., Hoadly 128–129.
Generally, a habitual drunkard could be assigned a guardian or committed to an asylum only after proceedings before a probate court or something like it. See, e.g., Ball & Roane 456. And, typically, surety statutes required a proceeding before a justice of the peace or a comparable officer before a bond could be ordered (or, if a bond wasn’t posted, before a jail sentence could be imposed). E.g., Acts and Laws of the State of Conn. 189 (E.
Babcock ed. 1786); see also Rahimi, 602 U. S., at 699.
None of that holds true for §922(g)(3). On the government’s account, the statute automatically divests an individual of his constitutional right to bear arms the moment he becomes an unlawful user and until he ends his drug use—all without any pre-deprivation process. To be sure, and as the government highlights, an unlawful user who violates this ban is entitled to “a full-dress criminal trial” before he can be imprisoned or permanently disarmed.
Brief for United States 26. But be that as it may, §922(g)(3) offers an unlawful user no pre-deprivation process before his “‘temporary disarmament,’” the very burden the government says is akin to the burden vagrancy, civil-commitment, and surety laws imposed on habitual drunkards.
Ibid. —————— 6Certain other provisions of §922, such as subsections (g)(1) (disarming convicted felons) and (g)(4) (disarming any person “adjudicated as a mental defective” or “committed to a mental institution”), involve some manner of pre-deprivation process before an individual’s Second Amendment rights are lost. For that reason, they differ from subsection (g)(3) and “nothing in our opinion should be taken to cast doubt” on them.
D
We see one more problem yet with the government’s submission. Recall its claim about §922(g)(3)’s “why.” The government argues that, consistent with historical laws, §922(g)(3)’s unlawful user provision disarms individuals who are, as a “category,” “violen[t]” and “unusually dangerous.” Brief for United States 4, 10–11. So far, we have spotted the government its assertion about §922(g)(3)’s purpose. But there are at least two reasons to doubt it has established even that much.
The first has to do with §922(g)(3)’s reliance on the CSA. The former does not define its own category of people to disarm. Instead, according to the government, it disarms anyone who regularly uses any drug found on any CSA schedule for something other than its “prescribed purpose.” Tr. of Oral Arg. 57. The CSA, in turn, was adopted to protect “the health and general welfare of the American people.” 21 U. S. C. §801(2). Drugs can be added to its schedules for a variety of reasons having little or nothing to do with their potential to induce violence—reasons that include “[t]he state of current scientific knowledge” about a substance, whether that substance “is an immediate precursor” to another controlled substance, and the risk to “public health.” §811(c). Without question, some unlawful users of controlled substances can pose a risk of violence. But, by defining its scope through the CSA—a statute animated by a variety of other concerns—it is far from obvious that 18 U. S. C. §922(g)(3) confines its reach to those who are categorically and unusually dangerous.
—————— District of Columbia v. Heller, 554 U. S. 570, 626 (2008). Likewise, our conclusion today should not be taken to suggest “that the Second Amendment prohibits the enactment of laws banning the possession of guns by categories of persons thought by a legislature to present a special danger of misuse.” United States v. Rahimi, 602 U. S. 680, 698 (2024). The problem in this case is simply that the historical evidence the government presents does not support the categorical restriction it urges. The second reason has to do with the government’s approach to the drug at issue here. As this case came to us, marijuana was listed on Schedule I—a schedule reserved for drugs with “a high potential for abuse” with “no currently accepted medical use.” 21 U. S. C. §812(b)(1). But after we heard oral argument, the government moved some marijuana products to Schedule III, 91 Fed. Reg. 22714 (2026), a schedule that applies to drugs with a lower potential for dependence and abuse and for which a “currently accepted medical use” exists, §812(b)(3). Years before that, too, the Department of Justice issued a memorandum directing federal prosecutors nationwide to curtail their enforcement efforts against marijuana users even while all marijuana products remained on Schedule I. Attorney General Memo (Aug. 29, 2013).
Seismic changes followed that memorandum. While marijuana use largely remained unlawful under federal law, the number of federal offenders sentenced for possession of marijuana dwindled. See United States Sentencing Commission, Interactive Data Analyzer. And most States responded by legalizing marijuana use to one degree or another as a matter of state law. See Nat. Conf. of State Legislatures, State Medical Cannabis Laws (June 27, 2025) (“Forty states, three territories and the District of Columbia” have legalized some marijuana use). As a result, some surveys suggest there now may be more adults in this country who regularly use marijuana than consume alcohol.
See, e.g., J. Caulkins, Changes in Self-Reported Cannabis Use in the United States from 1979 to 2022, 119 Addiction 1648 (2024) (finding, for the first time in 2022, more individuals who self-report daily or near-daily marijuana use than alcohol use). Whatever one thinks of these developments, the federal government has not just tolerated them; it helped fuel them. All of which leaves it awkwardly positioned to suggest that the millions of Americans who now regularly use marijuana are categorically and unusually dangerous.
In saying this much, we do not question that sometimes an individual’s unlawful use of marijuana (or any other controlled substance) may render him a danger to others. But, again, the government disclaims the need to show anything like that in this case. Instead, it asks us to conclude that anyone who regularly uses marijuana is categorically violent and dangerous without any further showing. All based on little more than its current say-so, one at odds with its own regulatory actions. And affording the government that kind of “broad power to designate any group as dangerous and thereby disqualify its members from having a gun” would risk allowing it to “quickly swallow” the Second Amendment. Kanter v. Barr, 919 F. 3d 437, 465 (CA7 2019) (Barrett, J., dissenting).
* In many respects, this case is a narrow one. We do not address efforts to ban addicts, see n. 2, supra, or those presently intoxicated, from possessing a firearm. We do not address other prophylactic laws Congress might adopt after determining that users of a particular drug pose a special risk of misusing firearms. See n. 6, supra. We do not address 18 U. S. C. §922(g)(1)’s provision disarming individuals convicted of felonies (often including drug-related ones). Ibid. We do not even address whether the government could bring a prosecution under §922(g)(3) accompanied by individualized proof that the defendant’s use of marijuana (or any other drug) renders him a danger to himself or others. Or proof that a certain drug always renders its users dangerous because of its potency or for some other reason. None of those issues is before us and we do not pass on them either way.
All that is before us is one, if surely ambitious, theory. The government maintains that it may automatically strip Mr. Hemani of his Second Amendment right to possess a firearm because he uses marijuana a few times a week.
More than that, because he possessed a gun despite this prohibition, the government insists it may imprison him for up to 15 years and disarm him for life. According to the government, none of this turns on how much marijuana Mr.
Hemani uses or what effect it has on him. It makes no difference either if he keeps a firearm only in his home for self- defense, never misuses a gun while intoxicated, and never poses a danger to himself or others as a result of his marijuana use. The only thing the government must show, it says, is that an individual like Mr. Hemani regularly uses any amount of any controlled substance.
To square that expansive theory with the Second Amendment, the government invites us to draw an analogy between its present regulation and historical laws addressing habitual drunkards. Those laws, the government contends, demonstrate a tradition of firearm regulation consistent with its effort to disarm any regular user of any controlled substance without any further showing. But the government’s analogy fails under every measure it asks us to consider: The historical laws on which it relies targeted different kinds of people, did so for different reasons, and operated in different ways. And faced with all these shortcomings in the government’s submission, we cannot say it has carried its conceded burden of showing its prosecution of Mr. Hemani complies with the Second Amendment.
The judgment of the Fifth Circuit is affirmed.
It is so ordered.
_________________ _________________ SUPREME COURT OF THE UNITED STATES No. 24–1234 UNITED STATES, PETITIONER v. ALI DANIAL HEMANI ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT [June 18, 2026]