Context: Malaria has been increasing globally, and epidemics tend to occur when weather conditions favor this vector-borne disease. Long-term meteorologic forecasting using El Niño Southern Oscillation (ENSO) may assist in anticipating epidemics and targeting scarce resources.

Objective: To determine whether malaria epidemics in Venezuela are related to ENSO and rainfall and to determine whether such a relationship could be used to predict outbreaks.

Design: Retrospective analysis of national malaria morbidity (1975-1995) and mortality (1910-1935) data in the coastal zone and interior of Venezuela in relation to El Niño events and rainfall.

Main outcome measure: Correlation between malaria mortality and morbidity and sea surface temperatures (SSTs) in the Eastern Tropical Pacific, a parameter of ENSO.

Results: Malaria mortality and morbidity have increased by an average of 36.5% (95% confidence interval, 3.7%-69.3%; P=.004) in years following recognized El Niño events. A moderate correlation was found between Pacific tropical SST during a Niño event and malaria 1 year later (r=0.50, P<.001). Malaria mortality is more strongly related to drought in the year preceding outbreaks than to rainfall during epidemic years.

Conclusions: Historic and recent data from Venezuela demonstrate that malaria increases by an average of about one third in the year following a Nino event; change in malaria risk can be predicted from Pacific SSTs in the previous year. Therefore, the occurrence of an El Niño event may help predict malaria epidemics in this part of South America.