Several variables determine the impact of a pesticide onaquatic invertebrates. In this study, aquatic insects weresubjected to the common insecticide diazinon; we analyzed the variables taxon, developmental stage, and exposure time. Effects of diazinon on the caddis fly Hydropsyche angustipennis and the midge Chironomus riparius were determined in the laboratory during different exposure times (48 and 96 h) using mortality, activity, and growth as end points. Last instars of both species displayed a clear behavioral response at concentrations much lower than those affecting survival. Doubling the exposure time from 2 to 4 d decreased survival of midges and caddis flies by a factor 1.4 to 8.4. The 96-h 50% lethal concentrations were: 1.3 μg/L (first instar of the caddis fly), 29 μg/L (fifth instar of the caddis fly), 23 μg/L (first instar of the midge) and 167 μg/L (fourth instar of the midge). Within the spectrum of tested insects (nine species for which 48-h 50% lethal concentrations have been reported in the literature), H. angustipennis is the second most sensitive, and C. riparius the most tolerant species. However, the ranking of species strongly depends on the developmental stage; differences between species are often smaller than differences between instars of one species. The large difference in sensitivities between young and old larvae imply that the impact of a pesticide strongly depends on the season of occurrence. Runoff from pesticide applications on crops is more likely to occur in spring and summer and may have a relatively greater impact on insect communities since young larvae prevail in these seasons. In addition, recovery of typical riverine insects such as H. angustipennis from incidental exposure will be slow, considering their relatively long life cycle.