ORDER
The opinion filed March 2, 2011, is replaced by the amended opinion filed concurrently with this order.
With these amendments, Judges Ripple and Graber have voted to deny the petition for panel rehearing, and Judge Pregerson has voted to grant it. Judge Graber has voted to deny the petition for rehearing en banc, and Judge Ripple has so recommended. Judge Pregerson has voted to grant it.
The full court has been advised of the petition for rehearing en banc, and no judge of the court has requested a vote on it.
The petition for panel rehearing and petition for rehearing en banc are DENIED. No further petitions for rehearing will be accepted.
OPINION
Plaintiff Santiago Lopez appeals from the entry of summary judgment in favor of Defendant Pacific Maritime Association on Plaintiffs claims for disparate treatment and disparate impact under the federal Americans with Disabilities Act of 1990 (“ADA”) and the state Fair Employment and Housing Act (“FEHA”). On de novo review, Doran v. 7-Eleven, Inc., 524 F.3d 1034, 1047 (9th Cir.2008), we affirm.
I. Background1
Defendant represents the shipping lines, stevedore companies, and terminal operators that run the ports along the west coast of the United States. As the collective bargaining agent and payroll administrator of those employers, Defendant enforces the policies that govern the hiring of longshore workers who work along the west coast. One of those policies is a “one-strike rule,” which eliminates from consideration any applicant who tests positive for drug or alcohol use during the pre-employment screening process. Defendant notifies its applicants at least seven days in advance of administering the drug test. Failing the drug test, even once, disquali*764fíes an applicant permanently from future employment.
Plaintiff wants to be a longshoreman. He first applied in 1997 at the port in Long Beach, California. At that time, however, Plaintiff suffered from an addiction to drugs and alcohol. When Defendant administered its standard drug test, Plaintiff tested positive for marijuana. Defendant therefore disqualified Plaintiff from further consideration under the one-strike rule.
In late 2002, Plaintiff recognized the deleterious effects on his health that his addictions had caused. He became clean and sober and, in 2004, reapplied to be a longshoreman. Because of the one-strike rule, Defendant rejected Plaintiffs application. At that time, Defendant did not know of Plaintiffs earlier addiction. Plaintiff attempted to appeal, but Defendant never entertains appeals from disqualifications arising from positive drug tests.
Plaintiff then filed this action, claiming that Defendant violated the ADA and the FEHA by discriminating against him on the basis of his protected status as a rehabilitated drug addict. See 42 U.S.C. §§ 12112(a), 12114(b)(1) (expressly protecting from employment discrimination any person who has “been rehabilitated successfully and is no longer engaging in [illegal drug use]”); Cal. Gov’t Code § 12926(i) (incorporating by reference any protection in the ADA not expressly incorporated into the FEHA). The district court granted summary judgment to Defendant. Plaintiff timely appeals.
II. Plaintiffs Disparate Treatment Claim
In support of his disparate treatment claim, Plaintiff makes three arguments. He first argues that the one-strike rule facially discriminates against recovering or recovered drug addicts. We disagree. The rule eliminates all candidates who test positive for drug use, whether they test positive because of a disabling drug addiction or because of an untimely decision to try drugs for the first time, recreationally, on the day before the drug test. Conversely, the rule allows a drug-addicted applicant who happens to be sober at the time of the drug test to complete pre-employment processing successfully. Here, for example, had Plaintiff applied for the first time in 2004, he would qualify for employment despite his status as a recovering addict. The triggering event for purposes of the one-strike rule is a failed drug test, not an applicant’s drug addiction.
The Supreme Court’s opinion in Raytheon Co. v. Hernandez, 540 U.S. 44, 124 S.Ct. 513, 157 L.Ed.2d 357 (2003), supports our view of the one-strike rule. There, the Court held that an employer’s policy not to rehire workers who lost their jobs due to drug-related misconduct constituted a “neutral,” “legitimate, [and] nondiseriminatory reason for refusing to rehire” the aggrieved employee. Id. at 53, 124 S.Ct. 513. The Court disapproved of the argument that, because the employee’s misconduct related to his drug addiction, the employer’s refusal to rehire him on account of that misconduct violated the ADA. Id. at 54 n. 6, 124 S.Ct. 513. The ADA prohibits employment decisions made because of a person’s qualifying disability, not decisions made because of factors merely related to a person’s disability. See Hazen Paper Co. v. Biggins, 507 U.S. 604, 611, 113 S.Ct. 1701, 123 L.Ed.2d 338 (1993) (holding that an employer’s decision “wholly motivated by factors other than age” does not constitute age discrimination, even if “the motivating factor is correlated with age”).
Second, Plaintiff alleges that Defendant adopted the one-strike rule intentionally to exclude recovering and recovered drug ad-*765diets from its work force. The record belies that allegation. Before the adoption of the one-strike rule, the longshore industry suffered numerous serious accidents and injuries, including several fatalities. Defendant attributed those accidents in part to a culture that accepted the use of drugs and alcohol in the workplace. Defendant thought that it could reduce such accidents by eliminating applicants who might be more likely to use drugs or alcohol at work.
With the support of the union that represented the longshore workers, Defendant began administering drug tests to new applicants. Defendant decided to make the disqualification of applicants who tested positive permanent because it thought that applicants who could not abstain from using an illegal drug, even after receiving advance notice of an upcoming drug test, showed less responsibility and less interest in the job than applicants who passed the drug test. Thus, Defendant’s reasons for rejecting applicants who test positive did not include a calculation that an applicant might test positive because of a drug addiction, rather than because of recreational use.
In short, nothing about the history of the one-strike rule leads us to conclude that Defendant adopted the rule with a discriminatory purpose. The ADA and the FEHA protect people who are recovering or who have recovered from a drug addiction; they do not protect people who are using illegal drugs when they apply for a job. It was lawful for Defendant to eliminate applicants who were using drugs when they applied to be longshore workers. It was likewise lawful for Defendant to disqualify those applicants permanently. Nothing in the record suggests that Defendant targeted or attempted to target recovered drug addicts, as distinct from recreational users.
Finally, Plaintiff argues that the district court improperly granted summary judgment because Defendant learned of Plaintiffs drug addiction shortly after disqualifying him. We fail to see how Plaintiffs attempt to inform Defendant of his status after Defendant disqualified him has any bearing on whether Defendant decided to disqualify Plaintiff because o/his protected status. See Raytheon, 540 U.S. at 54 n. 7, 124 S.Ct. 513 (“If [the employer] were truly unaware that ... a disability existed, it would be impossible for her hiring decision to have been based, even in part, on [the employee’s] disability. And, if no part of the hiring decision turned on [the employee’s] status as disabled, he cannot, ipso facto, have been subject to disparate treatment.’-’); Brundage v. Hahn, 57 Cal. App.4th 228, 66 Cal.Rptr.2d 830, 836 (1997) (affirming summary judgment on an employee’s ADA and FEHA disparate treatment claims because the employer could not have discriminated on the basis of a disability of which it did not know).
Plaintiff attempts to rely on our opinion in Hernandez v. Hughes Missile Systems Co., 362 F.3d 564 (9th Cir.2004), to show that he has raised a genuine issue of material fact with respect to his disparate treatment claim. But the facts in Hernandez clearly distinguish it from the circumstances here. In Hernandez, an employer fired an employee for failing to pass his drug' test. Id. at 566. The employee later became sober and reapplied. Unlike in this case, the employee included with his new application a letter from his Alcoholics Anonymous sponsor that informed the employer of the employee’s “steady and consistent progress in his recovery from [his] disease.” Id. That piece of evidence was crucial because, even though the employer claimed that it had not read the letter, a reasonable juror could have found that it did. Id. at 566-67. Consequently, we reversed summary judgment because “a rea*766sonable jury could determine that [the employer] refused to rehire [the employee] because of his past record of addiction and not because of a company rule barring rehire of previously terminated employees.” Id. at 570.
Plaintiff has no such evidence. His letter informing Defendant of his drug addiction came only after Defendant’s decision to disqualify him from further processing. Accordingly, Plaintiffs disparate treatment claim fails because there is no evidence in this record that Defendant disqualified him because of his protected status.
III. Plaintiff’s Disparate Impact Claim
Plaintiffs disparate impact claim also fails. He argues that the one-strike rule disparately affects recovering drug addicts by eliminating anyone who previously tested positive for drug use. But he offers only the bald assertion that this result must be so. We disagree because, as we have noted, the rule does not necessarily screen out recovering drug addicts disproportionately.
A. Initially, we observe that Plaintiff has litigated his disparate impact claim from start to finish on the familiar theory taken from discrimination claims brought under Title VII, 42 U.S.C. §§ 2000e to 2000e-4. In Plaintiffs opposition to Defendant’s motion for summary judgment, for example, he told the district court that, “[i]n order to state a prima facie case of disparate impact discrimination, the Plaintiff must show that employment practices that are facially neutral fall more harshly on one group than another and cannot be justified by business necessity.” He cited two Title VII cases, Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988), and Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). His opening brief to us restated the same principle, again citing Watson as controlling precedent. The district court concluded from that argument that Plaintiff intended to prove his disparate impact claim through the usual method of proving “disparate impact” as set out in Title VII cases. We did, too.
In a petition for rehearing filed after our original opinion issued, however, Plaintiff changed his theory. With support from several amici curiae,2 Plaintiff now argues that we erred by using Title VII cases to guide our analysis, because disparate impact claims under the ADA require a different approach. Specifically, § 12112(b)(6) of the ADA makes unlawful any “selection criteria that screen out or tend to screen out an individual with a disability or a class of individuals with disabilities unless ... consistent with business necessity.” Applying § 12112(b)(6) here, Plaintiff argues, he need not make a “formal statistical showing” of a disparate impact because the one-strike rule tends to screen out recovered drug addicts, whether it does so in fact or not.
Plaintiff plainly waived his new theory, and we therefore do not address it on *767the merits.3 To allow Plaintiff to make this argument now, after the case has been litigated, appealed, briefed, submitted, and decided, would deprive us of the assistance of our colleague below and would deprive Defendant of the opportunity to meet Plaintiffs new argument in the proper course. For those reasons, we adhere to the twin requirements that a litigant must argue clearly all of his theories of relief, both in the district court and in his opening brief, to preserve those theories on appeal. Alameda Books, Inc. v. City of Los Angeles, 631 F.3d 1031, 1044 (9th Cir. 2011); United States v. City of Arcata, 629 F.3d 986, 992 (9th Cir.2010). Because Plaintiff failed to raise this new argument in either of those places, he has waived- it. We emphasize that we express no view on what application, if any, § 12112(b)(6) would have to the facts of this case. That question is not before us.
B. To create a genuine issue of fact under the theory that Plaintiff chose to pursue, Plaintiff must have produced evidence from which a fact-finder reasonably could conclude that the one-strike rule results in fewer recovered drug addicts in Defendant’s employ, as compared to the number of qualified recovered drug addicts in the relevant labor market. See Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 308, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977) (comparing the racial composition of the employer’s work force to the racial composition of the qualified work force in the relevant labor' market). The record contains neither statistical nor anecdotal evidence to that effect.
Plaintiff did present to the district court an affidavit from a forensic economist. The economist concluded that Plaintiff had a viable disparate impact claim4 because, after “[comparing the selection rate of the whole population [of longshore worker applicants] (48%) versus the population of the ‘protected group’ (0%),” she found it “evident that [the one-strike rule] is in violation of the 80% Rule.” See 29 C.F.R. § 1607.4(D) (“A selection rate' for any race, sex, or ethnic group which is less than [80%] of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a'greater than [80%] rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact.”).
Although the economist did not define the “protected group,” she appears to have made the same mistake as Plaintiff. The correctly defined “protected group” for purposes of the ADA and the FEHA is recovered drug addicts. It does not matter whether such individuals applied to Defendant previously. If the number of recovered addicts in Defendant’s workforce roughly reflects the number of recovered addicts in the relevant labor market, then Defendant has not broken the law under Plaintiffs theory. By defining the “protected group” as the number of recovered drug addicts who previously applied to Defendant and were rejected because they failed a drug test, Plaintiff and his expert assume their own conclusion. Although Defendant will never hire those *768people and thus their selection rate always will be zero, we still do not know how many recovered drug addicts Defendant hires versus how many recovered drug addicts it turns away, nor do we know how many of those turned away are not drug addicts, recovering or otherwise.
Plaintiff complains that our standard places an unfair burden on him because he has no way to know how many recovering or recovered drug addicts Defendant has disqualified. Neither can he determine the proportion of recovering or recovered drug addicts in the relevant labor market because, he argues, state law prevents him from inquiring into a person’s history of drug abuse. We recognize the challenge involved in bringing a disparate impact claim of this kind, but both logic and precedent require him to produce some evidence that tends to show that the one-strike rule excludes recovering or recovered drug addicts disproportionately. Plaintiff introduced no such evidence. At the summary judgment stage, a party no longer can rely on allegations alone, however plausible they may be. Berger v. City of Seattle, 569 F.3d 1029,1077-78 (9th Cir.2009) (en banc). Because the record contains no evidence supporting an essential element of Plaintiffs disparate impact claim, we must affirm the summary judgment on that claim.
IV. Conclusion
We recognize that the one-strike rule imposes a harsh penalty on applicants who test positive for drug use. As Defendant candidly concedes, many people question the rule’s reasonableness in light of the fact that many people who use drugs later rehabilitate themselves, as Plaintiff exemplifies. But unreasonable rules do not necessarily violate the ADA or the FEHA. Because Plaintiff failed to establish that Defendant intentionally discriminated against him on the basis of his protected status or that the one-strike rule disparately affects recovered drug addicts, we affirm the summary judgment in favor of Defendant.
AFFIRMED.