An outdoor artificial stream system was used to examine the effects of a chronic application of the organophosphate pesticide chlorpyrifos on the invertebrate fauna of the system. Two replicate streams received chlorpyrifos for 21 days at the high dose (5 micrograms.liter-1) or low dose (0.1 microgram.liter-1) or only the carrier solution with which chlorpyrifos is mixed for commercial sale (1,1,1-trichloroethane and xylene). Streams behaved as replicates with respect to five water quality parameters. Seventy-four nonchironomid and 24 chironomid taxa were recorded during the study. The number of taxa and total invertebrate abundance were significantly reduced by both high and low doses of the pesticide. Shannon-Weaver diversity was also reduced by both high and low doses of the pesticide whereas evenness increased in high dose streams. The individual abundances of 9/36 nonchironomid and 13/19 chironomid taxa were significantly reduced by pesticide application; the abundance of one taxon, the gastropod Physastra, increased. The biomass of periphyton in the streams was affected by changes in the abundance of chironomid grazers and Physastra, and the slow recovery of grazers from mortality due to chlorpyrifos appeared to result in a higher biomass of periphyton in the high dose streams than in controls and low dose streams 21 days after dosing ceased. Inverse relationships between the amounts of fine particulate organic matter (FPOM) and the sum of the numbers of all taxa of collector-gatherers (as well as numbers of four individual taxa) were interpreted as disruption of the processing of FPOM by members of this trophic group following toxic effects of chlorpyrifos. The implications of the study for biological monitoring of the direct and indirect effects of chronic doses of chlorpyrifos on streams are discussed, and the use of chironomid larvae and periphyton in biomonitoring to detect pesticide impacts is particularly recommended.