A study was undertaken in Debre Zeit sector, central Ethiopia to identify the most important determinants of malaria transmission with a specific purpose of assessing whether global warming was the main cause of increased morbidity and mortality ascribed to malaria in the highlands of Ethiopia. Both retrospective and prospective methods were employed to conduct the study in 430 localities with a total population of 406,891. Some nine data sets were collected including altitude, malaria incidence and prevalence, hospital morbidity and mortality, outbreaks of malaria, vector control, climate patterns and in-vivo drug resistance mostly on a monthly basis with varying time periods ranging from 1951 to 1993. Morbidity analysis revealed a 67-fold increase in monthly incidence of malaria in about two decades. Mortality patterns showed a 13-fold increase in deaths ascribed to malaria in the last decade alone. Furthermore, highland communities living in localities lying between 2,000 and 2,200 metres were affected by P. falciparum transmission for the first time since 1986. Time series analysis of climate patterns revealed a trend of increased climatic warming in both day-time and night-time temperature especially since 1988, at which time a coincident peak in the incidence rate of malaria was also observed despite a decrease in total rainfall. Each °C rise in monthly mean night-time temperature was associated with up to 64% and 58% estimated rise in monthly incidence of falciparum malaria two and three months later respectively. The historic epidemic of malaria in which 150,000 people were estimated to have died among 3 million cases of malaria in Ethiopia in 1958 was associated with abnormally high ambient temperature and rainfall. This year also saw a very strong El Niño event. A simultaneous peak was seen in incidence of malaria, hospital admissions and hospital deaths ascribed to malaria together with an abnormal rise in mean night-time temperature in 1988. Based on current data it is concluded that epidemics of malaria in the highlands of Ethiopia that were observed during the past decade were mainly due to an increase in night-time temperature. The coincident peak in both malaria and ambient temperature together with statistical evidence suggested that global warming was the main cause of the rise in incidence of highland malaria. This appeared to be the cause of new foci of transmission at high altitude localities while also increasing both the rate and duration of transmission in previously known epidemic-prone areas from highly seasonal to perennial transmission. Furthermore, non-climatic biological and human conditions such as chloroquine resistant falciparum malaria, decreased vector control efforts and large scale population migration were also identified as important factors amplifying the impact of global warming on morbidity and mortality ascribed to highland malaria by affecting more distal parts of the causal pathway.