Lassa virus infection and antibodies were studied in households where Lassa fever cases occurred, and compared to those in nearest neighbor houses and "far" houses located across the village from case houses. Seventy-nine percent of all rodents caught in the houses were Mastomys, the natural reservoir of Lassa virus. Rodent infection was not randomly distributed, but rather focal. Thirty-nine percent of the Mastomys in case houses were viremic, compared to 3.7% in control houses. Human antibody prevalence in case houses was 30%, compared to 20% in non-case houses (P less than 0.05, chi-square test, df = 2). Neither seroconversions nor antibody prevalence rates were associated with household size or number of persons per room. Trapping of rodents in half of the case and control houses resulted in a Mastomys reduction ranging from 2.2- to 3.3-fold. This reduction failed to significantly reduce the seroconversion rate to Lassa virus in the people of trapped houses compared to those in untrapped ones. More complete trapping will be needed in order to better evaluate this procedure as a means of interruption of Lassa virus transmission in endemic villages.