Field and laboratory toxicity tests were conducted on the grass shrimp, Palaemonetes pugio, to evaluate the usefulness of laboratory testing in estimating mortality from fenvalerate exposure associated with agricultural runoff. The study examined an integrated approach for assessing the impacts of fenvalerate on estuarine fauna, using 96-h static-renewal and 6-h pulsed-dose laboratory toxicity tests and in situ toxicity tests. The laboratory toxicity tests with fenvalerate gave 96-h LC50 values ranging from 0.007 to 0.071 microgram/L and 6-h PDLC50 values ranging from 0.100 to 0.130 microgram/L. Comparisons of the results of two field toxicity tests with laboratory-derived LC50 values showed good agreement between field and laboratory toxicity data. The variation between field and laboratory toxicity tests may have been due to the limitations of the water sampling regime used in characterizing the pesticide exposure during the field toxicity tests. These comparisons suggest that a combination of laboratory and field toxicity testing is required to estimate the actual field mortality from fenvalerate exposure associated with agricultural runoff. Future studies should include composite water sampling and more frequent discrete sampling methods to better characterize field exposure regimes. (Copyright (c) 1989 SETAC.)