BackgroundMalaria vector intervention and control programs require reliable and accurate information about vector abundance and their seasonal distribution. The availability of reliable information on the spatial and temporal productivity of larval vector habitats can improve targeting of larval control interventions and our understanding of local malaria transmission and epidemics. The main objective of this study was to evaluate two methods of estimating larval habitat productivity in the western Kenyan highlands, the aerial sampler and the emergence trap.MethodsThe study was conducted during the dry and rainy seasons in 2008, 2009 and 2010. Aerial samplers and emergence traps were set up for sixty days in each season in three habitat types: drainage ditches, natural swamps, and abandoned goldmines. Aerial samplers and emergence traps were set up in eleven places in each habitat type. The success of each in estimating habitat productivity was assessed according to method, habitat type, and season. The effect of other factors including algae cover, grass cover, habitat depth and width, and habitat water volume on species productivity was analysed using stepwise logistic regressionResultsHabitat productivity estimates obtained by the two sampling methods differed significantly for all species except for An. implexus. For for An. gambiae s.l. and An. funestus, aerial samplers performed better, 21.5 and 14.6 folds, than emergence trap respectively, while the emergence trap was shown to be more efficient for culicine species. Seasonality had a significant influence on the productivity of all species monitored. Dry season was most productive season. Overall, drainage ditches had significantly higher productivity in all seasons compared to other habitat types. Algae cover, debris, chlorophyll-a, and habitat depth and size had significant influence with respect to species.ConclusionThese findings suggest that the aerial sampler is the better of the two methods for estimating the productivity of An. gambiae s.l. and An. funestus in the western Kenya highlands and possibly other malaria endemic parts of Africa. This method has proven to be a useful tool for monitoring malaria vector populations and for control program design, and provides useful means for determining the most suitable sites for targeted interventions.