Fish may affect macrobenthic communities through trophic interactions and modification of habitat conditions. We compared the density, diversity and taxonomic composition of larval chironomids under low vs high fish impact, created by stocking semi-natural ponds with either young/small common carp Cyprinus carpio at low biomass densities or with high biomass densities of older/large common carp, supplemented by other cyprinids and predatory fish species. Over the study season, May-August, total chironomid abundance was considerably lower in ponds with high fish biomass than in low fish biomass ponds. In July-August larval densities declined significantly, irrespective of the fish status of the ponds. The composition of larval communities diverged between ponds with different fish status. Chironomus riparius and C. plumosus accounted for most of the observed dissimilarity. C. plumosus densities were not affected by the fish status of the pond, constituting on average> 38 % of chironomid specimens in both types of ponds. C. riparius was abundant in ponds with low fish biomass (on average 30 %), but rare in ponds with high fish biomass (4 %). Other common taxa were Glyptotendipes pallens (15 % and 13 %, respectively) and Procladius sp. (14 % in ponds with low fish impact). Canonical correspondence analysis showed that fish biomass was more important in determining chironomid community composition than environmental variables indicative of pond eutrophication (total N, P, chlorophyll a and conductivity). Despite adverse effects on total abundance, chironomid diversity was higher under apparently stronger trophic pressure by fish. Fish may have relieved some chironomid species from invertebrate predation and competition with dominant Chironomus larvae. Other important drivers of chironomid assemblage and diversity patterns may include species-specific feeding modes and prey size selectivity of fish, the ability of chironomid larvae to attain size- and depth-refuge from small fish but not larger fish, and differential vulnerability of freeliving and tube-dwelling larvae relative to ontogenetic niche shifts of fish.