The larvae of most benthic marine invertebrate species must develop for a minimum of time in the plankton before they become competent for settlement and metamorphosis in response to stimulating external cues. In an experimental laboratory study, we identified the temporal window of cue receptivity within the moulting cycle of the megalopa stage of an estuarine crab, Chasmagnathus granulata Dana. This species shows an export strategy including an early larval transport to coastal marine waters where zoeal development takes place, followed by the return of the megalopa stage to brackish habitats where the adults live. In two series of experiments (A, B), megalopae were exposed for differential periods to a combination of metamorphosis-stimulating cues which had previously been found effective (seawater conditioned with adult odor and presence of mud). In experimental series A, these cues were added on successively later days of the moulting cycle, while series B comprised treatments in which the cues were provided from the first day (postmoult) and removed on successively later days of the moulting cycle. Each series of experiments was repeated with larvae originating from two different females (F1, F2). The average development time of megalopae kept continuously in the presence of cues (control experiments, C1) ranged in the two hatches from 9.3 to 9.6 days. In the inverse controls where no cue was added at any time (C2), megalopal development to metamorphosis took on average 11.2-12.0 days. In series A, development duration in treatments with exposure to the cues commencing within 3-4 days after moulting was not significantly different from that in the permanently exposed C1 controls. Later beginning of the exposure, by contrast, had no stimulating effect (significant delay compared to C1, no significant difference from unexposed control, C2). In series B, no significant differences in development time were observed between the C1 controls and treatments with an initial exposure for a minimum of 4 or 6 days of the moulting cycle (F1, F2, respectively). Shorter initial periods of exposure had no metamorphosis-stimulating effects (no significant difference from C2). In conclusion, our results from both experiments suggest that the megalopa stage of C. granulata is most receptive of stimulating cues during a period lasting from ca. one third to one half of the moulting cycle, which coincides with the transition between stages C (intermoult) and D0 (early premoult) of Drach's classification system. This suggests an interaction of extrinsic stimulating cues with intrinsic (hormonal) factors involved in the control of the moulting cycle.