The times of onset and completion of the hunting flights of three colonies of neotropical bats, each comprising 100–200 individuals, were observed for nine months. The colonies were of different species: Molossus ater (M.a.) and Molossus molossus (M.m.) of the Molossidae, and Myotis nigricans (My. n.) of the Vespertilionidae. Individuals of Phyllostomus hastatus (P.h., Phyllostomidae) were also observed. All the bats roosted in a building near Restrepo, Colombia (4°16′N, 73°34′W). Times of emergence in the evening and the return of the last animals in the morning were recorded on 2 to 3 successive days each month. For all bats, the emergence time changed in parallel with that of sunset, and the return paralleled sunrise (Fig. 1). Accordingly, the duration of the activity period is positively correlated with the duration of the night. No annual periodic changes in phase (re sunset/sunrise) of the onset and end of flight activity could be demonstrated, but there was a close relationship between the timing of activity and particular light intensities during twilight (Fig. 4). The first flyers of M.a. appear at the highest intensity (30–300 lx) and those of My. n. at the lowest (0.1–5 lx); the last flyers to return appear in the opposite sequence. For each species, the return to the roost usually occurs at a lower intensity than the departure. These findings, made with four neotropical bat species, differ from those of Subbaraj and Chandrashekaran (1977) with the emballonurid bat Taphozous that they studied at 9°58′ N in India. The ecological factors that may play a role in timing the flight activity of tropical bats are discussed. “Sunset-related timing”, based on the combined effect of (a) the circadian oscillation in “arousal” and (b) the transition during twilight to a light-intensity range with reduced inhibition of activity (lightsampling behavior), tends to be the rule in tropical bats; “time-of-day-related timing” is the exception.