dissenting.
The Court does not dispute the District Court’s critical findings of fact: First, Roundup Ready Alfalfa (RRA) can contaminate other plants. See App. to Pet. for Cert. 38a, 54a, 62a. Second, even planting in a controlled setting had led to contamination in some instances. See id., at 69a-70a. Third, the Animal and Plant Health Inspection Service (APHIS) has limited ability to monitor or enforce limitations on planting. See id., at 70a. And fourth, genetic contamination from RRA could decimate farmers’ livelihoods and the American alfalfa market for years to come. See id., at 71a; see also id., at 29a-30a. Instead, the majority faults the District Court for “enjoining APHIS from partially deregulating RRA.” Ante, at 158.
*167In my view, the District Court may not have actually ordered such relief, and we should not so readily assume that it did. Regardless, the District Court did not abuse its discretion when, after considering the voluminous record and making the aforementioned findings, it issued the order now before us.
I
To understand the District Court’s judgment, it is necessary to understand the background of this litigation. Petitioner Monsanto Company (Monsanto) is a large corporation that has long produced a weedkiller called Roundup. After years of experimentation, Monsanto and copetitioner Forage Genetics International (FGI) genetically engineered a mutation in the alfalfa genome that makes the plant immune to Roundup. Monsanto and FGI’s new product, RRA, is “the first crop that has been engineered to resist a[n] herbicide” and that can transmit the genetically engineered gene to other plants. See App. to Pet. for Cert. 45a.
In 2004, in the midst of a deregulatory trend in thé agricultural sector, petitioners asked APHIS to deregulate RRA, thereby allowing it to be sold and planted nationwide. Id., at 101a. Rather than conducting a detailed analysis and preparing an “environmental impact statement” (EIS), as required by the National Environmental Policy Act of 1969 (NEPA) for every “major Federal actio[n] significantly affecting the quality of the human environment,” 42 U. S. C. § 4332(2)(C), APHIS merely conducted an abbreviated “environmental assessment” (EA). During the 6-month period in which APHIS allowed public comment on its EA, the agency received 663 comments, 520 of which opposed deregulation. App. to Pet. for Cert. 29a. Farmers and scientists opined that RRA could contaminate alfalfa that has not been genetically modified, destroying the American export market for alfalfa and, potentially, contaminating other plants and breeding a new type of pesticide-resistant weed. Id., at 29a-30a.
*168Despite substantial evidence that RRA genes could transfer to other plants, APHIS issued a “Finding of No Significant Impact” and agreed to deregulate RRA “unconditionally,” ante, at 146. With no EIS to wait for and no regulation blocking its path, petitioners began selling RRA. Farmers and environmental groups swiftly brought this lawsuit to challenge APHIS’s decision to deregulate, raising claims under NEPA and other statutes.
The District Court carefully reviewed a long record and found that “APHIS’s reasons for concluding” that the risks of genetic contamination are low were “not ‘convincing.’” App. to Pet. for Cert. 38a. A review of APHIS’s internal documents showed that individuals within the agency warned that contamination might occur. APHIS rested its decision to deregulate on its assertion that contamination risk is “not significant because it is the organic and conventional farmers’ responsibility” to protect themselves and the environment. Ibid. Yet the agency drew this conclusion without having investigated whether such farmers “can, in fact, protect their crops from contamination.” Ibid. The District Court likewise found that APHIS’s reasons for disregarding the risk of pesticide-resistant weeds were speculative and “not convincing.” Id., at 46a. The agency had merely explained that if weeds acquire Roundup resistance, farmers can use “ ‘[alternative herbicides.’ ” Ibid. In light of the “acknowledged” risk of RRA gene transmission and the potential “impact on the development of Roundup resistant weeds,” the court concluded that there was a significant possibility of serious environmental harm, and granted summary judgment for the plaintiffs. Id., at 54a; see also id., at 45a.
At this point, the question of remedy arose. The parties submitted proposed final judgments, and several corporations with an interest in RRA, including Monsanto, sought permission to intervene. The District Court granted their motion and agreed “to give them the opportunity to present *169evidence to assist the court in fashioning the appropriate scope of whatever relief is granted.” Id., at 54a (internal quotation marks omitted).
While the District Court considered the proposed judgments, it issued a preliminary injunction. Ordinarily, the court explained, the remedy for failure to conduct an EIS is to vacate the permit that was unlawfully given — the result of which, in this case, would be to prohibit any use of RRA. See id., at 55a; see also id., at 65a. But this case presented a special difficulty: Following APHIS’s unlawful deregulation order, some farmers had begun planting genetically modified RRA. Id., at 55a. In its preliminary injunction, the District Court ordered that no new RRA could be planted until APHIS completed the EIS or the court determined that some other relief was appropriate. But, so as to protect these farmers, the court declined to prohibit them from “harvesting, using, or selling” any crops they had already planted. Id., at 56a. And “to minimize the harm to those growers who intend to imminently plant [RRA],” the court permitted “[t]hose growers who intend to plant [RRA] in the next three weeks and have already purchased the seed” to go ahead and plant. Id., at 58a (emphasis deleted). Essentially, the court grandfathered in those farmers who had relied, in good faith, on APHIS’s actions.
Before determining the scope of its final judgment, the District Court invited the parties and intervenors to submit “whatever additional evidence” they “wish[ed] to provide,” and it scheduled additional oral argument. Id., at 58a-59a. The parties submitted “competing proposals for permanent injunctive relief.” Id., at 60a. The plaintiffs requested that no one — not even the grandfathered-in farmers — be allowed to plant, grow, or harvest RRA until the full EIS had been prepared. Id., at 64a. APHIS and the intervenors instead sought a remedy that would “facilitate] the continued and dramatic growth” of RRA: a “partial deregulation” order that would permit planting- subject to certain conditions, *170such as specified minimum distances between RRA and conventional alfalfa and special cleaning requirements for equipment used on the genetically modified crop. See id,., at 60a-64a.
The court adopted a compromise. First, it declined to adopt the APHIS-Monsanto proposal. APHIS itself had acknowledged that “gene transmission could and had occurred,” and that RRA “could result in the development of Roundup-resistant weeds.” Id., at 61a-62a. In light of the substantial record evidence of these risks, the court would not agree to a nationwide planting scheme “without the benefit of the development of all the relevant data,” as well as public comment about whether contamination could be controlled. Id., at 68a. The “partial deregulation” proposed by petitioners, the court noted, was really “deregulation with certain conditions,” id., at 69a — which, for the same reasons given in the court’s earlier order, requires an EIS, ibid. The court pointed out numerous problems with the APHISMonsanto proposal. Neither APHIS nor Monsanto had provided “evidence that suggests whether, and to what extent, the proposed interim conditions” would actually “be followed,” and comparable conditions had failed to prevent contamination in certain limited settings. Id., at 69a-70a. APHIS, moreover, conceded that “it does not have the resources to inspect” the RRA that had already been planted, and so could not possibly be expected “to adequately monitor the more than one million acres of [RRA] intervenors estimate [would] be planted” under their proposal. Id., at 70a. That was especially problematic because any plan to limit, contamination depended on rules about harvesting, and farmers were unlikely to follow those rules. Id., at 71a. “APHIS ha[d] still not made any inquiry” into numerous factual concerns raised by the court in its summary judgment order issued several months earlier. Id., at 70a.
Next, the court rejected the plaintiffs’ proposed remedy of “enjoin[ing] the harvesting and sale of already planted” *171RRA. Id., at 76a. Although any planting or harvesting of RRA poses a contamination risk, the court reasoned that the equities were different for those farmers who had already invested time and money planting RRA in good-faith reliance on APHIS’s deregulation order. And small amounts of harvesting could be more easily monitored. Rather than force the farmers to tear up their crops, the court imposed a variety of conditions on the crops’ handling and distribution. Id., at 77a.
As to all other RRA, however, the court sided with the plaintiffs and enjoined planting during the pendency of the EIS. Balancing the equities, the court explained that the risk of harm was great. “[Contamination cannot be undone; it will destroy the crops of those farmers who do not sell genetically modified alfalfa.” Id., at 71a. And because those crops “cannot be replanted for two to four years,” that loss will be even greater. AncL On the other side of the balance, the court recognized that some farmers may wish to switch to genetically modified alfalfa immediately, and some companies like Monsanto want to start selling it to them just as fast. But, the court noted, RRA is a small percentage of those companies’ overall business; unsold seed can be stored; and the companies “‘have [no] cause to claim surprise’” as to any loss of anticipated revenue, as they “were aware of plaintiffs’ lawsuit” and “nonetheless chose to market” RRA. Id., at 72a.
Thus, the District Court stated that it would “vacat[e] the June 2005 deregulation decision”; “enjoi[n] the planting of [RRA] in the United States after March 30, 2007,” the date of the decision, “pending the government’s completion of the EIS and decision on the deregulation petition”; and impose “conditions on the handling and identification of already-planted [RRA].” Id., at 79a. On the same day, the court issued its judgment. In relevant part, -the judgment states:
“The federal defendants’ June 14, 2005 Determination of Nonregulated Status for [RRA] is VACATED. Be*172fore granting Monsanto’s deregulation petition, even in part, the federal defendants shall prepare an [EIS]. Until the federal defendants prepare the EIS and decide the deregulation petition, no [RRA] may be planted.. ..
“[RRA already] planted before March 30, 2007 may be grown, harvested and sold subject to the following conditions.” Id., at 108a-109a.
II
Before proceeding to address the Court’s opinion on its own terms, it is important to note that I have reservations about the validity of those terms. The Court today rests not only the bulk of its analysis but also the primary basis for our jurisdiction on the premise that the District Court enjoined APHIS from partially deregulating RRA in any sense. See ante, at 152-153, 158-164.1 That is a permissible, but not necessarily correct, reading of the District Court’s judgment.
So far as I can tell, until petitioners’ reply brief, neither petitioners nor the Government submitted to us that the District Court had exceeded its authority in this manner. And, indeed, the Government had not raised this issue in any court at all. Petitioners did not raise the issue in any of their three questions presented or in the body of their petition for a writ of certiorari. And they did not raise the issue in their opening briefs to this Court. Only after re*173spondents alleged that Monsanto’s injury would not be redressed by vacating the injunction, insofar as RRA would still be a regulated article, did petitioners bring the issue to the Court’s attention. Explaining why they have a redress-able injury, petitioners alleged that the District Court’s order prevents APHIS from “implementing] an[y] interim solution allowing continued planting.” Reply Brief for Petitioners 5. APHIS, the party that the Court says was wrongly “barred . . . from pursuing any deregulation,” even “in accordance with the procedures established by law,” ante, at 161, did not complain about this aspect of the District Court’s order even in its reply brief.
Thus, notwithstanding that petitioners “adequately preserved their objection that the vacated deregulation decision should have been replaced by APHIS’s proposed injunction,” ante, at 150 (emphasis added), the key legal premise on which the Court decides this case was never adequately presented. Of course, this is not standard — or sound — judicial practice. See Kumho Tire Co. v. Carmichael, 526 U. S. 137, 159 (1999) (Stevens, J., concurring in part and dissenting in part). Today’s decision illustrates why, for it is quite unclear whether the Court’s premise is correct, and the Court has put itself in the position of deciding legal issues without the aid of briefing.
In my view, the District Court’s judgment can fairly be read to address only (1) total deregulation orders of the kind that spawned this lawsuit, and (2) the particular partial deregulation order proposed to the court by APHIS. This interpretation of the judgment is more consistent with the District Court’s accompanying opinion, which concluded by stating that the court “will enter a final judgment” “ordering the government to prepare an EIS before [the court] makes a decision on Monsanto’s deregulation petition.” App. to Pet. for Cert. 79a. The language of that opinion does not appear to “ba[r] the agency from pursuing any deregulation — no matter how limited,” ante, at 161 (emphasis deleted). *174This interpretation is also more consistent with APHIS’s own decision not to contest what, according to the Court, was an unprecedented infringement on the agency’s statutory authority.
To be sure, the District Court’s judgment is somewhat opaque. But it is troubling that we may be asserting jurisdiction and deciding a highly fáctbound case based on nothing more than a misunderstanding. It is also troubling that we may be making law without -adequate briefing on the critical questions we are passing' upon. I would not be surprised if on remand the District Court merely clarified its order.
HI
Even assuming that the majority has correctly interpreted the District Court’s judgment, I do not agree that we should reverse the District Court.
At the outset, it is important to observe that when a district court is faced with an unlawful agency action, a set of parties who have relied on that ¡action, and a prayer for relief to avoid irreparable harm, thej court is operating under its powers of equity. In such a case, a court’s function is “to do equity and to mould each decjree to the necessities of the particular case.” Hecht Co. v. Bowles, 321 U. S. 321, 329 (1944). “Flexibility” and “practicality” are the touchstones of these remedial determinations, as “the public interest,” “private needs,” and “competiiig private claims” must all be weighed and reconciled againstithe background of the court’s own limitations and its particular familiarity with the case. Id., at 329-330.2 I
*175When a district court takes on the equitable role of adjusting legal obligations, we review the remedy it crafts for abuse of discretion. “[D]eference,” we have explained, “is the hallmark of abuse-of-discretion review.” General Elec. Co. v. Joiner, 522 U. S. 136, 143 (1997). Although equitable remedies are “not left to a trial court’s ‘inclination,’ ” they are left to the court’s “‘judgment.’” Albemarle Paper Co. v. Moody, 422 U. S. 405, 416 (1975) (quoting United States v. Burr, 25 F. Cas. 30, 35 (No. 14,692d) (CC Va. 1807) (Marshall, C. J.)). The principles set forth in applicable federal statutes may inform that judgment. See United States v. Oakland Cannabis Buyers’ Cooperative, 532 U. S. 483, 497 (2001) (“[A] court sitting in equity cannot ignore the judgment of Congress, deliberately expressed in legislation” (internal quotation marks omitted)). And historically, courts have had particularly broad equitable power — and thus particularly broad discretion — to remedy public nuisances and other “ ‘purprestures upon public rights and properties,’ ” Mugler v. Kansas, 123 U. S. 623, 672 (1887),3 which include environmental harms.4
In my view, the District Court did not “unreasonably exercis[e]” its discretion, Bennett v. Bennett, 208 U. S. 505, 512 (1908), even if it did categorically prohibit partial deregulation pending completion of the EIS. Rather, the District Court’s judgment can be understood as either of two reasonable exercises of its equitable powers.
*176 Equitable Application of Administrative Law
First, the District Court’s decision can be understood as an equitable application of administrative law. Faced with two different deregulation proposals, the District Court appears to have vacated the deregulation that had already occurred, made clear that NEPA requires an EIS for any future deregulation of RRA, and partially stayed the vacatur to the extent it affects farmers who had already planted RRA.5
Under NEPA, an agency must prepare an EIS for “every . . . major Federal actio[n] significantly affecting the quality of the human environment.” 42 U. S. C. §4332(2)(C). Recall that the District Court had found, on the basis of substantial evidence, that planting RRA can cause genetic contamination of other crops, planting in controlled settings had led to contamination, APHIS is unable to monitor or enforce limitations on planting, and genetic contamination could decimate the American alfalfa market. In light of that evidence, the court may well have concluded that any deregulation of RRA, even in a “limited ... geographic area” with “stringent . . . regulations governing harvesting and distribution,”6 *177ante, at 161, requires an EIS under NEPA. See generally D. Mandelker, NEPA Law and Litigation §§8:83-8:48 (2d ed. 2009) (describing when an EIS is required); cf. Marsh v. Oregon Natural Resources Council, 490 U. S. 360, 371 (1989) (NEPA embodies “sweeping commitment” to environmental safety and principle that “the agency will not act on incomplete information, only to regret its decision after it is too late to correct”). Indeed, it appears that any deregulation of a genetically modified, herbicide-resistant crop that can transfer its genes to other organisms and cannot effectively be monitored easily fits the criteria for when an EIS is required.7 That is especially so when, as in this case, the environmental threat is novel. See Winter v. Natural Resources Defense Council, Inc., 555 U. S. 7, 23 (2008) (EIS is more important when party “is conducting a new *178type of activity with completely unknown effects on the environment”).8
Moreover, given that APHIS had already been ordered to conduct an EIS on deregulation of RRA, the court could have reasonably feared that partial deregulation would undermine the agency’s eventual decision. Courts confronted with NEPA violations regularly adopt interim measures to maintain the status quo, particularly if allowing agency action to go forward risks foreclosing alternative courses of action that the agency might have adopted following completion of an EIS. See Mandelker, NEPA Law and Litigation §4:61. The applicable regulations, to which the District Court owed deference,9 provide that during the preparation of an EIS, “no action concerning the [agency’s] proposal shall be taken which would . . . [h]ave an adverse environmental impact” or “[l]imit the choice of reasonable alternatives.” 40 CFR § 1506.1(a) (2009). As exemplified by the problem of what to do with farmers who had already purchased or planted RRA prior to the District Court’s judgment, even minimal deregulation can .limit future regulatory options. *179“Courts must remember that in many cases allowing an agency to proceed makes a mockery of the EIS process, converting it from analysis to rationalization.” Herrmann, Injunctions for NEPA Violations: Balancing the Equities, 59 U. Chi. L. Rev. 1263,1289 (1992); see also 40 CPR § 1502.5 (EIS should be implemented in manner ensuring it “will not be used to rationalize or justify decisions already made”).
Although the majority does not dispute that the District Court could have reasonably concluded that NEPA requires an EIS for even partial deregulation of RRA, it suggests that any such conclusion would have been incompatible with the court’s decision to permit limited harvesting by farmers who had already planted RRA. See ante, at 162.10 I do not see the “inconsistency].” Ibid. NEPA does not apply to actions by federal courts. See 40 CFR § 1508.12. Exercising its equitable discretion to balance the interests of the parties and the public, the District Court would have been well within its rights to find that NEPA requires an EIS before the agency grants “Monsanto’s deregulation petition, even in part,” App. to Pet. for Cert. 108a, yet also to find that a partial stay of the vacatur was appropriate to protect the interests of those farmers who had already acted in good-faith reliance on APHIS.
Similarly, I do not agree that the District Court’s ruling was “premature” because APHIS had not yet effected any partial deregulations, ante, at 160. Although it is “for the agency to decide whether and to what extent” it will pursue deregulation, ante, at 159, the court’s application of NEPA to APHIS’s regulation of RRA might have controlled any deregulation during the pendency of the EIS. Petitioners and APHIS had already come back to the court with a proposed partial deregulation order which, the court explained, *180was incompatible with its determination that there is a substantial risk of gene spreading and that APHIS lacks monitoring capacity. That same concern would apply to any partial deregulation order. The court therefore had good reason to make it clear, upfront, that the parties should not continue to expend resources proposing such orders, instead of just moving ahead with an, EIS. Cf. Railroad Comm’n of Tex. v. Pullman Co., 312 U. S. 496, 500 (1941) (“The resources of equity are equal to an adjustment that will avoid the waste of a tentative decision”). Indeed, it was APHIS itself that “sought to ‘streamline’ ” the process. Ante, at 160.
Injunctive Relief
Second, the District Court’s judgment can be understood as a reasonable response to the nature of the risks posed by RRA. Separate and apart from NEPA’s requirement of an EIS, these risks were sufficiently serious, in my view, that the court’s injunction was a permissible exercise of its equitable authority.
The District Court found that gene transfer can and does occur, and that if it were to spread through open land the environmental and economic consequences would be devastating. Cf. Amoco Production Co. v. Gambell, 480 U. S. 531, 545 (1987) (“Environmental injury, by its nature, can seldom be adequately remedied by money damages and is often permanent or at least of long duration, i. e., irreparable”). Although “a mere possibility of a future nuisance will not support an injunction,” courts have never required proof “that the nuisance will occur”; rather, “it is sufficient. . . that the risk of its happening is greater than a reasonable man would incur.” 5 J. Pomeroy, A Treatise on Equity Jurisprudence and Equitable Remedies § 1937 (§ 523), p. 4398 (2d ed. 1919) (first emphasis added). Once gene transfer occurred in American fields, it “would be difficult — if not impossible — to reverse the harm.” Hollingsworth v. Perry, 558 U. S. 183, 195 (2010) (per curiam).
*181Additional considerations support the District Court’s judgment. It was clear to the court that APHIS had only limited capacity to monitor planted RRA, and some RRA had already been planted. The marginal threat posed by additional planting was therefore significant. Injunctive remedies are meant to achieve a “nice adjustment and reconciliation between the competing claims” of injury by “mouldpng] each decree to the necessities of the particular case.” Weinberger v. Romero-Barcelo, 456 U. S. 305, 312 (1982) (internal quotation marks omitted). Under these circumstances, it was not unreasonable for the court to conclude that the most equitable solution was to allocate the limited amount of potentially safe RRA to the farmers who had already planted that crop.11
The Court suggests that the injunction was nonetheless too sweeping because “a partial deregulation need not cause respondents any injury at all...; if the scope of the partial deregulation is sufficiently limited, the risk of gene flow to their crops could be virtually nonexistent. ” Ante, at 162-163. The Court appears to reach this conclusion by citing one particular study (in a voluminous record), rather than any findings of fact.12 Even assuming that this study is correct, the *182Court ignores the District Court’s findings that gene flow is likely and that APHIS has little ability to monitor any conditions imposed on a partial deregulation. Limits on planting or harvesting may operate fine in a laboratory setting, but the District Court concluded that many limits will not be followed and cannot be enforced in the real world.13
Against that background, it was perfectly reasonable to wait for an EIS. APHIS and petitioners argued to the District Court that partial deregulation could be safely implemented, they submitted evidence intended to show that planting restrictions would prevent the spread of the newly engineered gene, and they contested “virtually every factual issue relating to possible environmental harm.” Geertson Seed Farms v. Johanns, 570 F. 3d 1130, 1135 (CA9 2009). But lacking “the benefit of the development of all the relevant data,” App. to Pet. for Cert. 68a, the District Court did not find APHIS’s and petitioners’ assertions to be convincing. I cannot say that I would have found otherwise. It was reasonable for the court to conclude that planting could not go forward until more complete study, presented in an EIS, showed that the known problem of gene flow could, in reality, be prevented.14
*183The District Court’s decision that more study was needed to assess whether limits on deregulation could prevent environmental damage is further reinforced by the statutory context in which the issue arose. A court’s equitable discretion must be guided by “recognized, defined public policy.” Meredith v. Winter Haven, 320 U. S. 228, 235 (1943); see also Hecht Co., 321 U. S., at 331 (explaining that when a court evaluates an agency’s decision against the background of a federal statute, the court’s discretion “must be exercised in light of the large objectives of the Act”). Congress recognized in NEPA that complex environmental cases often require exceptionally sophisticated scientific determinations, and that agency decisions should not be made on the basis of “incomplete information.” Marsh, 490 U. S., at 371. Congress also recognized that agencies cannot fully weigh the consequences of these decisions without obtaining public comments through an EIS. See Robertson v. Methow Valley Citizens Council, 490 U. S. 332, 350 (1989).15 While a court may not presume that a NEPA violation requires an injunction, it may take into account the principles embodied in the statute in considering whether an injunction would be appropriate. This District Court had before it strong evidence that gene transmission was likely to occur and that limits on growing could not be enforced. It also had a large amount of highly detailed evidence about whether growing restrictions, even if enforced, can prevent transmission. That evidence called into question the agency’s own claims *184regarding the risks posed by partial deregulation. In enjoining partial deregulation until it had the benefit of an EIS to help parse the evidence, the court acted with exactly the sort of caution that Congress endorsed in NEPA.
Finally, it bears mention that the District Court’s experience with the case may have given it grounds for skepticism about the representations made by APHIS and petitioners. Sometimes “one judicial actor is better positioned than another to decide the issue in question.” Miller v. Fenton, 474 U. S. 104, 114 (1985). A “district court may have insights not conveyed by the record.” Pierce v. Underwood, 487 U. S. 552, 560 (1988). In this case, the agency had attempted to deregulate RRA without an EIS in spite of ample evidence of potential environmental harms. And when the court made clear that the agency had violated NEPA, the agency responded by seeking to “ ‘streamline’ ” the process, ante, at 160, submitting a deregulation proposal with Monsanto that suffered from some of the same legal and empirical holes as its initial plan to deregulate. Against that background, the court may have felt it especially prudent to wait for an EIS before concluding that APHIS could manage RRA’s threat to the environment.
* * *
The District Court in this case was put in an “unenviable position.” Ibid. In front of it was strong evidence that RRA poses a serious threat to the environment and to American business, and that limits on RRA deregulation might not be followed or enforced — and that even if they were, the newly engineered gene might nevertheless spread to other crops. Confronted with those disconcerting submissions, with APHIS’s unlawful deregulation decision, with a group of farmers who had staked their livelihoods on APHIS’s decision, and with a federal statute that prizes informed decisionmaking on matters that seriously affect the environment, *185the court did the best it could. In my view, the District Court was well within its discretion to order the remedy that the Court now reverses. Accordingly, I respectfully dissent.
See also ante, at 161 (“[T]he District Court barred the agency from pursuing any deregulation — no matter how limited the geographic area in which planting of RRA would be allowed, how great the isolation distances mandated between RRA fields and fields for growing non-genetieallyengineered alfalfa, how stringent the regulations governing harvesting and distribution, how robust the enforcement mechanisms available at the time of the decision, and — consequently—no matter how small the risk that the planting authorized under such conditions would adversely affect the environment in general and respondents in particular” (emphasis deleted)).
See also, e. g., Railroad Comm’n of Tex. v. Pullman Co., 312 U. S. 496, 500 (1941) (“The history of equity jurisdiction is the history of regard for public consequences.. . . There have been as many and as variegated applications of this supple principle as the situations that have brought it into play”); Seymour v. Freer, 8 Wall. 202, 218 (1869) (“[A] court of equity ha[s] unquestionable authority to apply its flexible and comprehensive jurisdiction in such manner as might be necessary to the right administra*175tion of justice between the parties”). Indeed, the very “ground of this jurisdiction” is a court’s “ability to give a more complete and perfect remedy.” 2 J. Story, Equity Jurisprudence § 924, p. 225 (M. Bigelow ed. 13th ed. 1886).
See Steelworkers v. United States, 361 U. S. 39, 60-61 (1959) (per curiam) (reviewing history of injunctions to prevent public nuisances).
See, e. g., Georgia v. Tennessee Copper Co., 206 U. S. 230 (1907) (air pollution); Arizona Copper Co. v. Gillespie, 230 U. S. 46, 56-57 (1913) (water pollution).
See Reply Brief for Federal Respondents 3. There is an ongoing debate about the role of equitable adjustments in administrative law. See, e. g., Levin, “Vacation” at Sea: Judicial Remedies and Equitable Discretion in Administrative Law, 53 Duke L. J. 291 (2003). The parties to this appeal and the majority assume that the District Court’s remedy was crafted under its equity powers, and I will do the same.
One of the many matters not briefed in this case is how limited a partial deregulation can be. It is not clear whether the sort of extremely limited “partial deregulations” envisioned by the Court, see ante, at 161-164, in which RRA is “deregulated” in one small geographic area pursuant to stringent restrictions, could be achieved only through “partial deregulation” actions, or whether they could also (or exclusively) be achieved through a more case-specific permit process. Under the applicable regulations, a regulated article may still be used subject to a permitting process. See 7 CFR §§340.0, 340.4 (2010). These permits “prescribe confinement conditions and standard operating procedures ... to maintain confinement of the genetically engineered organism.” Introduction of Or*177ganisms and Products Altered or Produced Through Genetic Engineering, 72 Fed. Reg. 39021, 39022 (2007) (hereinafter Introduction).
Ordinarily, “[o]nee an article has been deregulated, APHIS does not place any restrictions or requirements on its use.” Id., at 39023. As of 2007, APHIS had never — not once — granted partial approval of a petition for nonregulated status. USDA, Introduction of Genetically Engineered Organisms, Draft Environmental Impact Statement, July 2007, p. 11, online at http://www.aphis.usda.gov/brs/pdf/complete_eis.pdf (as visited June 18, 2010, and available in Clerk of Court’s case file). In 2007, APHIS began contemplating a “new system” to allow for the release and use of genetically modified organisms, for “special eases” in which there are risks “that could be mitigated with conditions to ensure safe commercial use.” Introduction 39024 (emphasis added).
See, e. g., 40 CFR § 1508.8 (2009) (determination whether an EIS is required turns on both “[d]irect effects” and “[i]ndireet effects,” and “ineludefs] those resulting from actions which may have both beneficial and detrimental effects, even if on balance the agency believes that the effect will be beneficial”); § 1508.27(b)(4) (determination whether an EIS is required turns on “[t]he degree to which the effects on the quality of the human environment are likely to be highly controversial”); § 1508.27(b)(5) (determination whether an EIS is required turns on “[t]he degree to which the possible effects on the quality of the human environment are highly uncertain or involve unique or unknown risks”).
The Court posits a hypothetical in which APHIS deregulates RRA limited to a remote area in which alfalfa is not grown, and issues an accompanying order “mandating isolation distances so great as to eliminate any appreciable risk of gene flow to the crops of conventional farmers who might someday choose to plant in the surrounding area.” Ante, at 163. At the outset, it is important to note the difference between a plausible hypothetical and a piece of fiction. At least as of 2007, APHIS had never granted partial approval of a petition for nonregulated status. See n. 6, swpra. And I doubt that it would choose to deregulate genetically modified alfalfa in a place where the growing conditions and sales networks for the product are so poor that no farmer already plants it. Moreover, the notion that this imagined deregulation would pose virtually no environmental risk ignores one of the District Court’s critical findings of fact: APHIS has very limited capacity to monitor its own restrictions. The agency could place all manner of constraints on its deregulation orders; they will have no effect unless they are enforced.
See Marsh v. Oregon Natural Resources Council, 490 U. S. 360, 372 (1989).
The Court states that the order permitted both harvesting and planting. But the court’s final judgment permitted only sale and harvesting of RRA planted before March 30, 2007, more than a month before the judgment. See App. to Pet. for Cert. 109a; see also id., at 79a.
As explained previously, I do not see the court’s broad injunction as “inconsistent,” ante, at 162, with its decision that farmers who had already planted RRA could harvest their crop. The equities are different for farmers who relied on the agency than for companies like Monsanto that developed an organism knowing it might be regulated; and APHIS could monitor only a limited amount of RRA.
The Court also hypothesizes a set of growing conditions that would isolate RRA from the plaintiffs in this case, even if not from other farmers. See ante, at 163. As already explained, these hypotheticals are rather unrealistic. See n. 8, supra. And, given that the plaintiffs include environmental organizations as well as farmer and consumer associations, it is hard to see how APHIS could so carefully isolate and protect their interests. In any event, because APHIS concedes that it cannot monitor such limits, rules that protect these or any other parties may be merely hortatory in practice. Moreover, although we have not squarely addressed the *182issue, in my view “[t]here is no general requirement that an injunction affect only the parties in the suit.” Bresgal v. Brock, 843 F. 2d 1163, 1169 (CA9 1987). To limit an injunction against a federal agency to the named plaintiffs “would only encourage numerous other” regulated entities “to file additional lawsuits in this and other federal jurisdictions.” Livestock Marketing Assn. v. United States Dept. of Agriculture, 207 F. Supp. 2d 992, 1007 (SD 2002), aff’d, 335 F. 3d 711, 726 (CA8 2003).
The majority notes that the District Court acknowledged, at a hearing several months before it issued the judgment, that a simple but slightly overinclusive remedy may be preferable to an elaborate set of planting conditions. See ante, at 163-164, n. 6. Quite right. As the District Court said to APHIS’s lawyer at that hearing, if the agency issues an elaborate set of precautions, “I don’t know how you even start to enforce it.” App. to Pet. for Cert. 190a-191a.
I suspect that if APHIS and petitioners had come back to the court with more convincing evidence prior to completing an EIS, and moved to *183modify the court’s order, the court would have done so. Indeed, the District Court showed a willingness to recalibrate its order when it amended its judgment just a few months after the judgment’s issuance in light of APHIS’s submission that certain requirements were impractical. See App. to Pet. for Cert. llla-114a.
Accordingly, while “NEPA itself does not mandate particular results,” it does mandate a particular process and embodies the principle that federal agencies should “carefully conside[r] detailed information” before incurring potential environmental harm. Robertson, 490 U. S., at 350, 349.