The mutant genotype at codon 76 of the pfcrt gene (T76) has been proposed as a molecular marker for surveillance of chloroquine (CQ)-resistant Plasmodium falciparum malaria but this proposal has not been validated by population-based surveys. In 1998-99, in 6 Ugandan sentinel sites, the prevalence of P. falciparum infections with the T76 genotype and the level of CQ use were measured by community surveys, and CQ resistance was determined by in-vivo tests on 6-59-month-old children with clinical malaria. The prevalence of T76 was not related to the overall clinical (early and late treatment failure: ETF + LTF; r = 0.14, P = 0.78) or parasitological (RI + RII + RIII; r = 0.17, P = 0.73) CQ resistance. However, the percentage of individuals carrying only infections with the T76 genotype (T76 alone) increased with increasing ETF (r = 0.76, P = 0.07) and type RIII parasitological failure (r = 0.69, P = 0.12). Similarly, the ratio between T76 and K76 (the wild type) prevalences (T76/K76) was strongly and positively correlated with ETF (r = 0.85, P = 0.03) and RIII (r = 0.82, P = 0.04). Moreover, T76 alone (r = 0.90, P = 0.01) as well as T76/K76 (r = 0.90, P = 0.01) significantly increased with increasing community CQ use. T76 alone and T76/K76 can be useful markers to estimate the ETF and RIII prevalence as well as the amount of CQ use in the community.