We investigated the hypothesis that the increased incidence of malignant melanoma at the Lawrence Livermore National Laboratory (LLNL) in California was associated with a difference in medical care received by its employees compared with that received by other residents of the same geographic area. From records of the Kaiser Permanente Medical Care Program, to which about half of the LLNL employees belonged, we confirmed that the incidence of melanoma at the laboratory was 3.2 times that among members served by the Kaiser Permanente medical center in Walnut Creek, a nearby community (95% CI, 1.7-6.0). Rates of biopsy showing junctional, compound, and dermal nevi in men were also higher in the LLNL employees than in the other health-plan members (relative risks, 3.2, 2.4, and 1.3, respectively). When LLNL employees without melanoma were compared with community controls, we found that the employees had substantially more skin biopsies. Although the excess number of skin biopsies existed among LLNL employees before publicity about the problem, this excess increased after the publicity. Exposure to some environmental agent(s) or personal risk factors may have caused clinically suspicious pigmented lesions that required LLNL employees to have more skin biopsies. On the other hand, awareness of the laboratory's excess melanoma incidence may have increased physicians' propensity to obtain biopsy specimens of pigmented lesions.