This is an unblinded, voluntary, open-label toxicity trial assessing combination antiretroviral therapy of zidovudine plus lamivudine and, when indicated, indinavir or nelfinavir as agents for post-exposure chemoprophylaxis for health care workers (HCWs) sustaining occupational exposures to human immunodeficiency virus (HIV). Indinavir or nelfinavir may be added when the exposure is especially severe, or drug-resistant HIV is suspected. The two secondary purposes of this study are: 1.) to assess the effect of combination chemoprophylaxis on the temporal sequence of the appearance of markers of HIV infection (should it occur) in an HCW taking chemoprophylaxis; and 2.) to describe the epidemiology of exposures to HIV for which combination chemoprophylaxis is elected in institutions participating in this study.Antiretroviral drugs are initiated as soon as possible for the exposed HCW, but no longer than 72 hours post-exposure. The total duration of prophylaxis remains at 28 days. Exposed HCWs are followed for a minimum of 12 months. After the baseline enrollment visit, follow-up visits occur at 2, 4, and 6 weeks, and 3, 6, and 12 months. Individuals are assessed for drug toxicity as well as for early signs of HIV infection.Since July 1997, nine health care workers elected chemoprophylaxis on the protocol (seven were started on three drugs, two took only two drugs). Subjective toxicities occurred in all health care workers and included nausea, fatigue, headache, dyspepsia, and/or myalgia. Of the nine, four HCWs completed 28 days (two were on three-drug therapy). Four discontinued drugs early because of subjective toxicities (all were on three-drug therapy), and one HCW discontinued only indinavir on day 14 due to headache and nausea, but completed 28 days on two drugs. Symptomatology has not correlated with any objective hematologic or biochemical toxicities. We have had no infections at NIH among study participants.