Objective: To establish whether HIV seroprevalence in Bujumbura is sta
ble or continuing to increase. Methods: HIV seroprevalence data among
pregnant women from 1986 were compared with comparable data from 1991-
1992. Results: HIV seroprevalence among antenatal clinic attendees at
three sites was 10.5, 28.0 and 11.9% in 1986, compared with 7.7, 25.6
and 12.4%, respectively, in 1991-1992. A weighted least squares analys
is showed significant differences in HIV seroprevalence between the di
fferent sites (chi2, 71.71; P> 0.0001), but no evidence of any differe
nces between the 1986 and the 1991-1992 prevalence levels (chi2, 0.51;
P=0.6). Conclusions: Bujumbura appears to be in the endemic phase of
the spread of HIV. The stable geographic variation among clinic popula
tions in Bujumbura suggests the need for focused interventions, and a
general need for surveillance data to be gathered from numerous sites
so to identify those with the highest incidence of HIV infection.