This paper describes two cases of nafcillin-induced platelet dysfunction, with positive rechallenge data for one patient. Nafcillin resulted in abnormal bleeding times in both patients and a clinically apparent bleeding episode in one of the cases. Platelet function tests were performed on one patient during the initial therapy and after rechallenge with nafcillin. Platelet aggregation showed abnormal responses to ADP, collagen, and epinephrine. Platelet count and morphology were normal. Nafcillin should be recognized as another antibiotic which causes platelet function abnormalities and clinical bleeding episodes. sites, but the hematocrit did not decrease. At this time, debridement with possible grafting was planned for the leg ulcers. The preoperative laboratory screen revealed a template bleeding time of greater than 25 min (normal, 6 to 9 min), and the procedure was cancelled. The partial thromboplastin time and platelet count were normal. Platelet morphology was also normal. The prothrombin time was 15.6 s with a control of 13.6 s and was corrected with normal plasma. Liver function tests were normal, although 1 week earlier mild elevations of serum glutamic oxalacetic transaminase and serum glutamic pyruvic transaminase had been present. The patient's medications included prednisone (40 mg daily), double-strength trimethoprim-sulfamethoxazole (twice daily), sulindac (Clinoril; 200 mg twice daily), amitriptyline (50 mg at bedtime), and nafcillin as described above. All medications except prednisone and nafcillin were discontinued, and the bleeding time was repeated at 48 and 72 h. Platelet function tests revealed abnormal aggregation (see Table 1). Surgery was deferred, nafcillin was discontinued, and the patient was discharged. A bleeding time repeated 9 days later was normal (6.5 min).The patient was readmitted 1 month later for surgical repair of the extensor tendons on the right hand. The only medication at that time was prednisone (30 mg daily). The platelet count, prothrombin time, and partial thromboplastin time were normal, and the bleeding time was 6.5 min. The tendon repair was accomplished without excessive bleeding, and consent of the patient to rechallenge her with nafcillin was obtained. During this period, fat-containing foods were not allowed after 10 p.m., to prevent lipemic blood samples. No blood samples were drawn other than those done for coagulation 59