POSNER, Circuit Judge.

In 1995, the City of Chicago administered a new written test to 26,000 applicants for jobs as firefighters. After grading the tests, the City placed the applicants in three categories, based on their scores: "well qualified," "qualified," and "not qualified." The plaintiffs (and the members of their class) are black applicants who were placed in the "qualified" category. Applicants were told the test results within days after January 26, 1996, when notices of the results were mailed to all the applicants. On that day the mayor had announced that the test scores were in, but that "after all our efforts to improve diversity [including racial], these test results are disappointing." There were no names in his public announcement.

The notices stated that applicants in the qualified category were unlikely to be hired because of the large number whose scores had placed them in the "well qualified" category, but that the applicants rated "qualified" would remain on the eligible list (since they had passed the test) for as long as the list was used. In fact, as the media reported the next day, the City expected to hire only about 600 of the 1,782 applicants in the "well qualified" category in the next three years, implying that no one in the "qualified" category would be hired.

The suit, now entering its second decade, charges that the test had a disparate impact on the black applicants (that is, disproportionately classified them as "qualified" rather than "well qualified") and was not a valid test of aptitude for firefighting. If these things are true, the basing of hiring decisions on the test violated Title VII of the Civil Rights Act of 1964. After protracted proceedings, the district judge ruled in favor of the plaintiffs and decreed injunctive relief.

The City argues that the suit is untimely. The plaintiffs were required, as a pre-requisite to being allowed to sue, to file a charge with the EEOC within 300 days after their claim accrued. 42 U.S.C. § 2000e-5(e)(1); Stepney v. Naperville School District 203, 392 F.3d 236 (7th Cir. 2004). They filed their charge on March 21, 1997, which was 420 days after the date on which notice of the results of the test had been sent them and probably 417 to 419 days after they received the notice. But it was within 300 days of the City's beginning to hire applicants from the "well qualified" list, and the district judge ruled that the suit was therefore timely because each time the City hired applicants in the "well qualified" group as determined on the basis of the January 1996 test results it committed a fresh violation of Title VII that may have harmed "qualified" applicants.

The plaintiffs acknowledge that in a "disparate treatment" case, that is, a case of intentional discrimination, the charging period begins when the discriminatory decision is made, e.g., Ledbetter v. Goodyear Tire Rubber Co., ___ U.S. ___, 127 S.Ct. 2162, 2168, 2172, 167 L.Ed.2d 982 (2007); Lorance v. AT T Technologies, Inc., 490 U.S. 900, 910-11, 109 S.Ct. 2261, 104 L.Ed.2d 961 (1989); Delaware State College v. Ricks, 449 U.S. 250, 101 S.Ct. 498, 66 L.Ed.2d 431 (1980); Huels v. Exxon Coal USA, Inc., 121 F.3d 1047, 1051 (7th Cir. 1997); Cox v. City of Memphis, 230 F.3d 199, 204-05 (6th Cir. 2000), rather than when it is executed. We have held that if the plaintiff does not learn of the decision until later, the limitations period begins to run then. Cada v. Baxter Healthcare Corp., 920 F.2d 446, 450 (7th Cir. 1990); see also Oshiver v. Levin, Fishbein, Sedran Berman, 38 F.3d 1380, 1385-86 and n. 5 (3d Cir. 1994). But Hamilton v. 1st Source Bank, 928 F.2d 86, 87-88 (4th Cir. 1990) (en banc), is to the contrary, and the question was left open by the Supreme Court in Ledbetter v. Goodyear Tire Rubber Co., supra, 127 S.Ct. at 2177 n. 10. It is of no moment in this case.

In the Ricks case a college denied a faculty member tenure but offered him a "terminal" one-year contract, which he accepted. The Supreme Court held that the statute of limitations began to run from the denial of tenure rather than from the plaintiffs termination at the end of the one-year period, since that termination was the automatic consequence of the fact that he had only a one-year contract, rather than being the consequence of some fresh act of discrimination. It is the same here. The hiring only of applicants classified "well qualified" was the automatic consequence of the test scores rather than the product of a fresh act of discrimination.

The plaintiffs do not quarrel with the proposition that "well qualified" applicants should be hired ahead of those who are merely "qualified." They argue that the test that sorted applicants into those categories was discriminatory. That discrimination was complete when the tests were scored and, especially in light of the mayor's public comment about them, was discovered when the applicants learned the results. It's not as if the City had divided applicants into "a white branch" and "a Negro branch" and fixed a higher qualifying score for the latter; for then a refusal to hire a black who scored higher than a white but below the qualifying score for blacks would be an unmediated act of discrimination. See Bazemore v. Friday, 478 U.S. 385, 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986) (per curiam); Anderson v. Zubieta, 180 F.3d 329, 335-36 (D.C. Cir. 1999); Brinkley-Obu v. Hughes Training, Inc., 36 F.3d 336, 346 (4th Cir. 1994). The refusal to hire would not be due, even in the first instance, to the policy of basing hiring on test scores, since, by hypothesis, some blacks would have had higher scores than some whites yet, purely because of the racial division, would not have been hired. This case is different because "well qualified" is not a racial category, though its racial composition may have been influenced by a discriminatory decision taken earlier.

In Beavers v. American Cast Iron Pipe Co., 975 F.2d 792, 796-800 (11th Cir. 1992), on which the plaintiffs heavily rely, the employer limited insurance coverage to employees' children who lived with their employee parent, and the charge was that this discriminated against male employees. The plaintiffs sued long after the policy was adopted but within 180 days (the limitations period applicable to them) of the denial of their claim for dependent coverage, and this was held to be timely because the allegedly discriminatory policy was the sole cause of the denial; there was no intervening neutral act, as in this case.

The distinction is a fine one (and it is arguable on which side of it the facts of Beavers fell) but it is the distinction that the Supreme Court has drawn. The plaintiffs argue that it does not apply to a disparate-impact case, but we cannot think why not. The difference between the two types of discrimination case is not fundamental. Disparate-impact analysis, much like the McDonnell Douglas method of establishing a prima facie case, involves the use of circumstantial evidence to create an inference of discrimination. "The concept of disparate impact was developed for the purpose of identifying discriminatory situations where, through inertia or insensitivity, companies were following policies that gratuitously — needlessly — although not necessarily deliberately, excluded black or female workers from equal employment opportunities. Often these were policies that had been adopted originally for discriminatory reasons and had not been changed when the employer ceased deliberately discriminating — if he had; for another way of looking at the disparate impact approach is that it is primarily intended to lighten the plaintiffs heavy burden of proving intentional discrimination after employers learned to cover their tracks." Finnegan v. Trans World Airlines, Inc., 967 F.2d 1161, 1164 (7th Cir. 1992) (citations omitted). So if a test or other method of screening applicants for employment bears more heavily on one protected group than on another, the burden shifts to the employer to show that the method is a rational method of selecting employees. 42 U.S.C. § 2000e-2(k); see Allen v. City of Chicago, 351 F.3d 306, 311-12 (7th Cir. 2003); El v. Southeastern Pennsylvania Transportation Authority, 479 F.3d 232, 240-41 (3d Cir. 2007); Meacham v. Knolls Atomic, Power Laboratory, 461 F.3d 134, 139 (2d Cir. 2006). If he cannot show this, his continuing to use the test suggests that his purpose in doing so may be discriminatory, although that need not be shown.

Why any of this should change the date on which the statute of limitations begins to run escapes us; and years ago, in Davidson v. Board of Governors, 920 F.2d 441, 445 (7th Cir. 1990), we held that it does not. An applicant who fails to meet the employer's standard is hurt not by a fresh act of discrimination, but as the automatic consequence of an earlier one — the adoption of the standard. See also Cox v. City of Memphis, supra, 230 F.3d at 204-05; Bronze Shields, Inc. v. New Jersey Department of Civil Service, 667 F.2d 1074, 1083-84 (3d Cir. 1981).

The Ninth Circuit reached a contrary result in Bouman v. Block, 940 F.2d 1211, 1221 9th Cir. 1991), but did so on the mistaken premise that until the plaintiff was not promoted she could not be "certain" that the use of the allegedly discriminatory eligibility list would have that consequence and until she was certain her claim would not accrue. As explained in Davidson, if a plaintiff cannot by exercise of reasonable diligence determine within the statutory period whether he has been injured by an unlawful practice, then even though his claim accrued when the practice was adopted the doctrine of equitable tolling will allow him to delay suing until he can collect the information he needs in order to be able to sue. 920 F.2d at 445. (The plaintiffs in this case argue equitable tolling, but unavailingly as we shall see.) "[W]hen there is only one wrongful act the claim accrues with the first injury." Palmer v. Board of Education of Community Unit School District 201-U, 46 F.3d 682, 686 (7th Cir. 1995). The first injury in this case was the classification of the black applicants as merely "qualified" on the basis of a. test that they contend was discriminatory.

The plaintiffs argue in the alternative that the City's violation of Title VII was a "continuing violation." The phrase does not mean what it seems to mean. Suppose that year after year for ten years your employer does not pay you the minimum wage. That is a continuing violation in an acceptable sense of the term in ordinary language, though "repetitive violation" would be more precise. But the recurrent nature of the defendant's conduct would not entitle you to wait until year 15 (assuming the statute of limitations was five years) and then sue not only for the wages you should have received in year 10 but also for the wages you should have received in years 1 through 9. The statute of limitations begins to run upon injury (or discovery of the injury) and is not restarted by subsequent injuries. Knight v. Columbus, 19 F.3d 579, 581 (11th Cir. 1994); Hendrix v. City of Yazoo City, 911 F.2d 1102, 1103 (5th Cir. 1990); cf. Klehr v. A.O. Smith Corp., 521 U.S. 179, 190, 117 S.Ct. 1984, 138 L.Ed.2d 373 (1997). That is the "first injury" rule.

The doctrine of continuing violation allows you to delay suing until a series of acts by a prospective defendant blossoms into a wrongful injury on which a suit can be based. Limestone Development Corp. v. Village of Lemont, 520 F.3d 797, 801 (7th Cir. 2008); Reese v. Ice Cream Specialties, Inc., 347 F.3d 1007, 1011-12 (7th Cir. 2003); Huckabay v. Moore, 142 F.3d 233, 239 (5th Cir. 1998); Glass v. Petro-Tex Chemical Corp., 757 F.2d 1554, 1561 (5th Cir. 1985). Despite its name, it is a doctrine about cumulative rather than continuing violation. A typical case is workplace harassment on grounds of sex. The first instance of a coworker's offensive words or actions may be too trivial to amount to actionable harassment, but if they continue they may eventually amount to an actionable pattern of harassing behavior. And then the entire series is actionable. E.g., National Railroad Passenger Corp. v. Morgan, 536 U.S. 101, 117, 122 S.Ct. 2061, 153 L.Ed.2d 106 (2002); DeClue v. Central Illinois Light Co., 223 F.3d 434, 435 (7th Cir. 2000); Galloway v. General Motors Service Parts Operations, 78 F.3d 1164, 1166 (7th Cir. 1996); Jensen v. Henderson, 315 F.3d 854, 859 (8th Cir. 2002); Rush v. Scott Specialty Gases, Inc., 113 F.3d 476, 482 (3d Cir. 1997). If each harassing act had to be considered in isolation, there might be no actionable claim, even when by virtue of the cumulative effect of the acts it was plain that the plaintiff had suffered unlawful harassment. There is nothing, of that sort here. The plaintiffs were injured, and their claim accrued, when they were placed in the "qualified" category of the hiring list on the basis of their score in the firefighters' test; for that categorization delayed indefinitely their being hired.

Extension of the "continuing violation" doctrine in the manner urged by the plaintiffs would have ludicrous consequences. The plaintiffs received notification of their "qualified" status in 1995; could they ten years later ask to be hired as firefighters and when turned down sue the City for violating Title VII because the reason for not hiring them was that were not in the "well qualified" part of the hiring list? The answer implied by the plaintiffs' argument is "yes."

The plaintiffs further argue that? even if their claim accrued in January 1996, the running of the statute of limitations was tolled (stopped) because they could not determine within 300 days whether they had a case. The City claimed that its. hiring test had been validated by an expert, but it was slow to produce the expert report for the plaintiffs to scrutinize.

The doctrine of equitable tolling allows a plaintiff additional time within which to sue (or meet some other deadline) if even diligent efforts on his part would not have enabled him to prepare and file his suit within the statutory period. E.g., Beamon v. Marshall Ilsley Trust Co., 411 F.3d 854, 860-61 (7th Cir. 2005); Coda v. Baxter Healthcare Corp., supra, 920 F.2d at 451; Chung v. United States Department of Justice, 333 F.3d 273, 278-30 (D.C. Cir. 2003); EEOC v. Kentucky State Police Department, 80 F.3d 1086, 1096 (6th Cir. 1996). The question is whether the plaintiffs in this case knew enough within 300 days of the announcement of the test results to file a charge with the EEOC. The deadline is short, but a charging party is not required to conduct a precomplaint investigation, Coda v. Baxter Healthcare Corp., supra, 920 F.2d at 452, as he would have to do if he were filing a suit. To impose such a requirement would frustrate "a remedial scheme in which lay-persons, rather than lawyers, are expected to initiate the process." Edelman v. Lynchburg College, 535 U.S. 106, 115, 122 S.Ct. 1145,152 L.Ed.2d 188 (2002), quoting EEOC v. Commercial Office Products Co., 486 U.S. 107, 124, 108 S.Ct. 1666, 100 L.Ed.2d 96 (1988). The EEOC is supposed to do the investigating. EEOC v. Shell Oil Co., 466 U.S. 54, 68, 104 S.Ct. 1621, 80 L.Ed.2d 41 (1984).

And even a precomplaint investigation need not inquire into possible defenses, such as the defense that an employment requirement having a discriminatory impact is a bona fide qualification for hiring. To file a suit, you need only have a prima facie case; you are not required to plead the nonapplicability of possible defenses. Gomez v. Toledo, 446 U.S. 635, 640, 100 S.Ct. 1920, 64 L.Ed.2d 572 (1980). "Complaints need not contain any information about defenses and may not be dismissed for that omission." Xechem, Inc. v. Bristol-Myers Squibb Co., 372 F.3d 899, 901 (7th Cir. 2004) (emphasis in original). See also Oakes v. United States, 400 F.3d 92, 98 (1st Cir. 2005). The information bearing on the existence of a meritorious defense is likely to be in the defendant's possession, or at least more readily accessible to him than to the plaintiff; relative access is one of the criteria for parceling out issues between the plaintiffs case and the defendant's case. Moreover, precomplaint investigation of possible defenses would often be to a great degree wasted motion, because a plaintiff cannot be certain which defenses the defendant will plead, and so he would end up investigating some defenses that turned out not to be pleaded.

The plaintiffs' lawyer admitted at argument, moreover, that his reason for not filing the charge within 300 days was not that he needed more time to be able to file such a charge but that he didn't think it necessary because he thought that the statute of limitations would not begin to run until the City began hiring applicants from the "well qualified" category on the list. That was a fatal mistake.

The judgment is reversed with directions to enter judgment for the defendant.

REVERSED.