Catches of Prostephanus truncatus (Horn) and Teretriosoma nigrescens Lewis from 205 pheromone trapping stations in Mexico were compared and related to climate and habitat. Prostephanus truncatus numbers differed between regions and over time with significant interaction between the two. An overall peak in captures occurred between August and November but this was not apparent in Veracruz and Chiapas. Catches of T. nigrescens also differed but effects were less distinct. The peak in P. truncatus catch was related to monthly rainfall and, with a time lag, minimum temperature in the drier, upland regions, but poorly matched in Chiapas and Veracruz. Prostephanus truncatus abundance was predicted from climatic and habitat data during the periods March to 3 July, and 31 July to November. A regression equation derived from temperature, humidity and rainfall, explained 82% of the variance in the P. truncatus catch in the later period. Local habitat made only a small addition to this model. The best correlate of T. nigrescens numbers was the catch of P. truncatus, but it was less abundant relative to P. truncatus as P. truncatus became more numerous and where annual r.h. was above 70%. Predicted abundance of P. truncatus from the regression model is compared with predictions from laboratory studies of reproductive potential under different climatic regimes. While laboratory studies predict high survival in hot, humid areas, the trap data showed higher catches in cooler, more temperate regions. Possible reasons for these differences and implications for the spread of P. truncatus in Africa are discussed.