A technical fungicide mixture, R6 and its components, cymoxanil (CYM) and cupric oxychloride (Cu-OCl), were tested by sea urchin bioassays (Paracentrotus lividus and Sphaerechinus granularis). A set of toxicity endpoints was evaluated including both lethal and sublethal effects with the following endpoints: (a) acute embryotoxicity, (b) developmental defects, (c) changes in sperm fertilization success, (d) transmissible damage from sperm to the offspring, and (e) cytogenetic abnormalities. Acute effects on developing embryos were observed as early (prehatch) mortality at R6 levels greater than or equal to 25 mug/ml. The pesticide mixture R6 was tested at realistic concentrations, ranging from 25 ng/ ml to 2.5 mug/ml, and the two components, CYM and Cu-OCl, were tested, either alone or in mixture, at concentrations equal to their levels in the corresponding R6 solutions. RIS was either dissolved in filtered seawater (water only, W-O), or spiked in "pristine" silt-clay sediment or soil samples before bioassays. Developmental toxicity of R6, following W-O dissolution, displayed a significant dose-related increase of larval malformations and differentiation arrest at concentrations of 750 ng/ml to 2.5 mug/ml both in P. lividus and in S. granularis larvae. Developmental toxicity was removed in spiked sediment up to R6 nominal levels (25 mug/ml), 10-fold above the embryotoxic RIS levels in W-O exposure. No significant developmental toxicity was exerted by CYM or Cu-OCl (W-O exposure) up to their concentrations equivalent to 2.5 mug/ml R6. The laboratory-prepared mixture of CYM and Cu-OCl, in the same concentration range, only resulted in minor effects, as larval retardation, suggesting the presence of toxic impurities (or additional components) in the R6 formulation. When sperm from either P. lividus or S. granularis were exposed to R6 before fertilization, a W-O exposure resulted in a dose-related decrease in fertilization of P. lividus sperm (up to 250 mug/ml R6), whereas S. granularis sperm underwent a significant increase of fertilization rate at the highest R6 nominal levels (up to 25 mug/ml). Equivalent CYM or Cu-OCl levels were ineffective on sperm fertilization success in both species. The offspring of S. granularis sperm exposed to 25 mug/ml R6 showed a significant increase in larval malformations, which were not detected in the offspring of R6-exposed P. lividus sperm. Again, CYM or Cu-OCl was unable to exert any transmissible damage from sperm to the offspring in either species. The present study raises the case of possible discrepancies between toxicity of a technical mixture and of its analytical-grade components, also providing evidence for a loss of pesticide toxicity following dispersion in an environmental matrix such as sediment or soil.