The risks of agranulocytosis and aplastic anemia in relation to the use of anti-infective drugs were estimated in a population-based case-control study conducted in Europe and Israel. Anti-infective drug use in the 2-week period before the onset of illness was compared between 251 patients admitted to hospital with agranulocytosis and 1271 controls hospitalized for reasons judged to be unrelated to previous use of anti-infective drugs. Anti-infectives significantly associated with agranulocytosis when used for at least 3 consecutive days were trimethoprim/sulfamethoxazole (relative risk, 12; 95% confidence interval, 3.9 to 40) and macrolides (infinity). The relative risk estimate for any use of sulfonamides without trimethoprim was elevated, but not statistically significant (3.6; 0.7 to 18). These estimates took confounding by various factors, in particular the use of other drugs, into account. The estimated excess risks of agranulocytosis attributable to the use of trimethoprim/sulfamethoxazole and macrolides in a 2-week period were 1.6 and 7.1 per million, respectively. Anti-infective use during the 29- through 180-day period before hospital admission was compared between 135 patients with aplastic anemia and 1410 controls. Although relative risk point estimates were elevated for trimethoprim/sulfonamides (2.1), other sulfonamides (2.9), and beta-lactams (1.5), none was statistically significant.