Malaria in Kenya continues to pose a great threat to health. Trials with sulfadoxine/ pyrimethamine (SP), currently recommended to replace chloroquine (CQ) in drug-resistant areas, were carried out between 1997 and 1999 in Plasmodium falciparum endemic and epidemic sites. In vitro assays showed that over 50% of isolates were resistant to CQ and SP and sensitive to both quinine and mefloquine. In Busia, an endemic site, efficacy of combination treatment with C Q and SP was compared with SP alone. The combination was more effective in fever resolution. SP and co-trimoxazole (sulfamethoxazole/trimethoprim) were compared in the Oyugis and Tiwi endemic sites and all outcomes were found to be equivalent. Codon changes in dhps and dhfr genes (on chromosomes 8 and 4) conferring in vitro resistance to sulfadoxine and pyrimethamine, predicted SP treatment failures in Mosoriot, an epidemic site, but age-related recovery seen in the endemic site Tiwi masked the influence of mutations in dhps. Resistance- related mutations in dhfr and dhps were mutually associated in both sites, which probably indicates drug selection acting on both gene-products. Mutations in pfmdr1 and pfert genes (chromosomes 7 and 5) associated with in vitro C Q resistance, were examined in the Mwea endemic site. The pfertThr76 mutation was found on presentation in all CQ-susceptible and resistant cases. The pfmdr1 mutations, Tyr86 and Tyrl246, were similarly present in most of the pre-treatment infections, but an increase in prevalence was seen in samples from recrudescent infections. Genotyping of parasite strains revealed that parasites from both endemic and epidemic areas were genetically diverse suggesting that the source of epidemic outbreaks was endemic sites of high transmission. Quantitation of parasite-specific RNA by nucleic acid sequence-based amplification (QT-NASBA) was evaluated for the detection of low malaria parasitaemias. The test maybe useful in monitoring therapeutic response to drugs.