The long-term response, including recovery, of aquatic macroinvertebrates to short-term insecticide exposure may be affected by the presence of uncontaminated refuges in the stressed ecosystem. Experimental ditches were used to study the influence of non-sprayed ditch sections regarding the ecotoxicological effects on and the recovery of macroinvertebrates following treatment with the insecticide lufenuron. The treatment regimes differed in the proportion of the ditch (0, 33, 67, and 100% of surface area) that was sprayed to reach a lufenuron concentration of 3 μ/L in the water column of the sprayed ditch section. The magnitude and duration of effects on macroinvertebrates, and on arthropods in particular, were higher when a larger proportion of the ditch was treated. Initially, more pronounced responses were observed for bivoltine and multivoltine insects and macrocrustaceans than for univoltine and semivoltine arthropods. Most macroinvertebrate arthropods showed delayed responses, with maximum treatment-related effects observed two to six weeks after lufenuron application. This latency of effects can be explained by the mode of action of lufenuron, involving inhibition of chitin synthesis, which affects arthropod molting and metamorphosis. The observed effects were short-lived only in those ditches where 33% of the surface area was sprayed. In the ditches where 67 and 100% of the surface area was sprayed, some insects and macrocrustaceans showed long-term effects. In the 100% sprayed ditches in particular, the treatment-related reduction in arthropods resulted in indirect effects, such as an increase in snails, and later in an increase in the ephemeropteran Cloeon dipterum, probably because of an increase in periphyton, and release from competition and predation. Effects that are most likely indirect also were observed for Oligochaeta, Hirudinea, and the flatworm Mesostoma sp.