Between 1984 and 1989, the onchocerciasis-vector control zone on the pilot area of San Vicente Pacaya, Guatemala, which had been subject to experimental control since 1979, was extended from 91.3 to 148.6 km2. Temephos was used as a larvicide against Simulium ochraceum s.l. the target species. As a new strategy, only breeding sites with water discharges of 0.1-10 litres/s were treated, every 2 weeks. This approach provided a substantial reduction in effort, number of treated sites, time and cost. To assess the effect of the temephos, nine sites were selected in which standardized collections of adult Simulium were made twice a month, by human bait. There was an obvious difference between the pre- and post-control mean densities of flies at each site. In the northern area, which includes the Lavaderos, Barretal, Colina and Rodeo sites, the biting density in 1979, before treatment, varied between 10 and 64 flies/man-hour (FMH). Four years later, this had been reduced to 0.1-3.2 FMH, and by the end of the present study, in 1989, the mean density was zero FMH. In the southern area, which lies south-east of Lavaderos (and includes Guachipilin, Ingerto, Pena Blanca and Sierra Morena), the density during the pre-control phase was 24 FMH at one of the two sites investigated at the time and 39.3 FMH at the other. It fell to 0.1-0.5 FMH after 5 years of control and to zero (three sites) or close to zero (< or = 0.5 FMH; one site) for the last 4 years of the present study. To assess the effect of vector control on onchocerciasis prevalence and incidence, 1280 residents from six endemic communities, out of 12,000 permanent inhabitants, were examined. In Santa Cruz, Patrocinio and Los Rios, the prevalence of skin microfilariae in the subjects from each community fell from 8.1%-37.8% during the pretreatment, base-line period to 0.0%-31.5% when the study foci were totally integrated into the vector-control operation following treatment. Incidence among children (aged < or = 9 years) varied from 0%-25% for the period 1982-1984 but, thereafter, not a single case appeared in four of the six study communities (Santa Cruz, Patrocinio, Los Rios and Berlin). Incidence in Guachipilin did not decline appreciably, probably because of human migration into the area from other onchocerciasis foci. The prevalence of nodules followed a similar trend to those of the prevalence and incidence of skin microfilariae, falling from 9.1%-45.0% pre-control to 1.8%-14.3% 10 years later.