Descriptions of primary HIV-1 infection have so far been based on Caucasians living in industrialized nations. Due to studies of leptospirosis in the predominantly black population of Barbados, serum was available for patients admitted with acute febrile illnesses to the Queen Elizabeth Hospital (QEH). By searching the medical records of 510 adult patients with known HIV-1 infection we identified 10 patients who had stored serum from an admission for an acute febrile illness that predated or coincided with their first HIV-1-positive test. Serological testing confirmed primary HIV-1 infection in 9 and was suggestive in the 10th patient. The clinical features of these 10 patients were in keeping with previous descriptions of primary HIV-1 infection but differed from leptospirosis cases seen at the QEH. One patient died during his seroconversion illness and another died 3 months after seroconversion. The findings suggest that severe primary HIV-1 infection could be a relatively uncommon occurrence, that the condition may be misdiagnosed, and that cases may not occur until the AIDS epidemic is established. :A retrospective review was conducted of the medical records of 510 HIV-1-positive adult patients who had attended the Queen Elizabeth Hospital (QEH) to determine whether any had been admitted for an illness compatible with a diagnosis of primary HIV-1 infection. A serum bank, created from patients who had been admitted with acute febrile illnesses and investigated for leptospirosis, provided serological evidence for primary HIV-1 infection in 10 patients. Serological testing of the serum samples confirmed primary HIV-1 infection in nine patients and was suggestive in the tenth. The clinical features of the 10 patients fit the earlier descriptions of primary HIV-1 infection, but differed from the leptospirosis cases seen at the QEH. One patient died during his seroconversion illness and another died 3 months after seroconversion. These findings suggest that severe primary HIV-1 infection could be a relatively uncommon occurrence, that the condition may be misdiagnosed, and that cases may not occur until the AIDS epidemic is established.