The time taken by Aedes aegypti to take a bloodmeal was determined under laboratory and field conditions in Trinidad. Eggs from field-collected females were reared under nutrient-limited (low diet) and nutrient-adequate (high diet) conditions. Blood-feeding times indicated that 71% of the females from the low-diet group were fast feeders (< 2 min) compared with 40% of the female from the high-diet group. Subsequent tests of the F1 and F2 progeny of the fast-feeding females indicated that the fast-feeding characteristic was not maintained. Similarly, testing of colonized Ae. aegypti (ST JOSEPH) indicated that, although 48% of the first, adult females were fast feeders (< 2 min), there was no apparent selection for the fast-feeding characteristic in mass-rearing of the F1 and F2 progeny from these females. Mean wing length (range = 1.66-2.94 cm) of the colonized females (2.67 cm) was more similar to that of females from the high-diet group (2.62 cm) than to that of females from the low-diet group (1.85 cm). Although wing lengths of females from landing collections in St Joseph, Trinidad, varied from 2.28-2.76 cm [mean (S.D.) = 2.46 (0.11) cm], there was no significant correlation between wing length and blood-feeding time on humans. The mean (S.D.) duration of blood-feeding for the wild Ae. aegypti, 179 (32) s, was comparable with that for females reared in the laboratory under nutrient-adequate conditions. Although the blood-feeding times for the wild females ranged from 107-282 s, only 2% (1 of 40) exhibited fast feeding times (< 2 min). Even though it is well known that mosquito size determines the probability of 'feeding success,' the duration of blood-feeding by wild Ae. aegypti on humans appeared independent of mosquito size (although the sample size was small). There may therefore be a genetic component to mosquito feeding speed, this trait being quickly lost during colonization and modulated by environmental factors.