A prospective cohort study of 121 HIV-1-positive homosexual men was conducted in Miami, Florida, U.S.A. to evaluate the associations between plasma zinc and copper levels and mortality. Plasma zinc and copper levels were measured at baseline and then at semiannual visits. Zinc inadequacy and copper inadequacy were defined as plasma zinc levels <75 (microg/dl) and plasma copper levels <85 (microg/dl), respectively. HIV-1-related deaths were confirmed by review of death certificates. Cox proportional hazards regression models with time-dependent covariates were used to estimate the relative risks of zinc and copper inadequacy on mortality. Over the average course of the 3.3-year follow-up, 19 participants (16%) died of HIV-1-related causes. After adjustment for potential confounders, including low CD4+ cell counts and antiretroviral therapy, zinc inadequacy and copper:zinc ratio >1 (i.e., plasma copper level greater than plasma zinc level) were associated with increased mortality (relative risks [RRs]; 95% confidence intervals [CIs], 4.98, 1.30-19.00 and 8.28, 1.03-66.58, respectively). A negative association was also observed between plasma zinc levels and mortality (RR 0.94; 95% CI, 0.91-0.98). Plasma levels of copper were not significantly associated with mortality. These results suggest that plasma zinc inadequacy or the plasma copper:zinc ratio may be useful predictors of survival in HIV-1 infection. The latter appears to be a stronger predictor.