We report here the results of a 42-day follow-up of 224 patients treated with mefloquine for uncomplicated falciparum malaria in Eastern Thailand. The risk of treatment failure among the 209 patients who completed the follow-up was 57% (95% confidence interval [CI] = 50,64). The daily risk of treatment failure, estimated through an exponential survival function, was constant during the follow-up period and equal to 1.9%. Five patient characteristics were found to be independent and important predictors of treatment failure in a Cox proportional hazards model: a young age (adjusted hazard ratio [AHR] for an increase of 10 years of age = 0.75 [95% CI = 0.62,0.90]); a history of at least three documented mefloquine treatments in the past year (AHR = 1.89 [95% CI = 1.10,3.24]); a parasitemia count > 100,000/µl (AHR = 1.80 [95% CI = 1.16,2.80]); and a history of diarrhea in the first two days after treatment (AHR = 1.51 [95% CI = 0.99,2.31]). A hemoglobin level ≤ 10 g/dl on the day of consultation was also a predictor of treatment failure, but only among patients who had been treated for malaria in the past 120 days (AHR = 4.38 [95% CI = 1.55,12.1]). These latter patients may have become anemic while they were unsuccessfully treated with mefloquine for an infection by a multiple drug-resistant strain in the past 120 days. It appears that the anemic patients may have been included in the study at the time of recrudescence from this original infection, and they were at higher risk of treatment failure since the strain infecting them was mefloquine-resistant. Children living in malarial areas, being at higher risk of treatment failure, may represent a suitable sentinel population for the monitoring of in vivo drug resistance.