The impact of biotech-soybean technology on freshwater ecosystems is being evaluated in the Rolling Pampas region, Argentina. The effect of cypermethrin, the main soybean insecticide, on low-order temperate-stream fish populations was investigated for two consecutive crop cycles under field-use conditions in biotech-soybean production. Cypermethrin was unable to induce mortality or behavioral effects on any of the fish species resident in a first-order stream across a crop field (pulsed acute exposure scenario) sprayed according to conventional practices. No spatially or temporally dependent effects were observed on population parameters (size-class structure, abundance, survival, sex and immature/mature ratio, condition factor) of resident or caged Cnesterodon decemmaculatus after spraying or rainfall events, not even one year after, at the beginning of the next crop cycle. Although cypermethrin was "very highly toxic" to C. decemmaculatus in laboratory water (96h-LC(50)=0.43microg/l), its toxicity was reduced in filtered (78%) and unfiltered (92%) stream water. Changes in LC(50) values were mainly correlated with the OC content of each water fraction (r(2)=0.99; p<0.01; n=9), showing that both DOC and TOC contributed proportionally to toxicity reduction. Protective effects of stream water (12-fold reduction LC(50) values) explained the lack of effects on fish populations in the field, despite cypermethrin water concentrations after spraying reached values comparable with the 96h-LC(50). Therefore, cypermethrin under field-use conditions in transgenic-soybean production represents a low risk of acute exposure for fish populations inhabiting low-order temperate-streams rich in TOC. The relationship between LC(50) and TOC could be a convenient way to improve risk estimation based on laboratory toxicity testing.