Abstract
The prevalence of antibodies to HIV in population groups in Dar es Salaam, Tanzania has been monitored from late 1984 through 1991. in the pregnant women and blood donors the doubling time was about 2 years. If, however, the doubling time is calculated from reported cases of AIDS of the whole country, an average doubling time of 10 months is obtained. We point out factors that may bias calculations of the rate of spread of the virus and thus give rise to erronous projections for the nearest years, especially the so-called transient effect. It is recommended that serosurveys of representative population groups are carried out at regular intervals.