Bacteriological isolation of Yersinia pestis is the reference test for confirming plague infection, but recovery of the pathogen from human samples is usually very poor. When the etiology of the disease cannot be bacteriologically confirmed, it may be useful to possess alternative tests such as detection of specific circulating antibodies to help guide the diagnosis. In the present study, the immunoglobulin G (IgG) anti-F1 enzyme-linked immunosorbent assay (ELISA) has been applied to various human sera to evaluate its large-scale applicability in the high-endemicity plague foci of Madagascar. The sensitivity of the test was found to be 91.4%, and its specificity was 98.5%. The positive and negative predictive values were 96 and 96.6%, respectively. Seroconversion was observed on day 7 after onset of the disease. Patients with a positive ELISA result could be separated into high (82%) and low (18%) IgG anti-F1 responders. Cross-reactions with eight other infectious diseases prevalent in Madagascar were scarce and were found in 1 of 27 Mycobacterium tuberculosis-, 3 of 34 Schistosoma haematobium-, and 1 of 41 Salmonella-infected patients. Finally, the efficiency of the IgG anti-F1 ELISA was evaluated during the Mahajanga, Madagascar, plague outbreak of 1995. When the number of ELISA-positive patients was added to the number of bacteriologically confirmed and probable cases, the number of positive patients was increased by 35%. In conclusion, although it does not replace bacteriology, IgG anti-F1 ELISA is a useful and powerful tool for retrospective diagnosis and epidemiological surveillance of plague outbreaks.