Many factors might be related to this situation. Lindblade (2000) suggested that malaria is unstable in highland areas. It can make the population in that area who have little or no immunity, vurneable to the explosion of malaria when the vector density and weather conditions favor transmission (Lindblade, 2000; MARA, 1998). In the areas over 850 m above sea level, the malaria incidence pattern from the median malaria incidence in 1995-2000 indicates that there were three peaks of malaria transmission which are in May, August and December. This is one of the indicators that malaria transmission in this area may be unstable.