A prospective study on the prevalence of HIV-l infection in Karachi, Pakistan was conducted over a period of six years (1986-1992). Over 15,000 individualsamples and more than 32,000 donor units of individuals residing in Karachi at the time of sample co llection were tested for HIV-I infection by our screening test EIA which revealed a positivity rate of 0.23% and 0.003% in individual and donor units respectively by Western Blot. We divided patients into four groups A,B,C and D based on the most plausible cause of transmission. The largest number of positive patients belonged to group B, who were of either foreign origin or expatriates or Pakistanis settled abroad. They comprised approximately 67% of the total positive cases and were subjected to testing on strong clinical grounds. In individuals of other groups like group A and D, there was history of travel abroad from time to time. The only positive donor unit (group C) belonged to a person who had been li ving in Middle East for the last 10-12 years. The last group D comprised of samples that wer e directly sent to us without complete history,except for the fact that they had been travelling back and forth. The large majority of patients fell in 20-50 years age group. Despite the limitations of this study, we conclude that the prevalence of HIV is steadily Increasing in our population and so far, we have no been able to find an indigenous case of AIDS in our series