Contrapsin, a serine protease inhibitor (serpin) present in mouse serum, was compared with that found in adult Schistosoma mansoni worm homogenates, which although immunologically identical to contrapsin in mouse serum, had a higher molecular weight in Western blotting. Immunolocalization studies demonstrated parasite-associated contrapsin on the surface and interstitial cells of adult male worms. After extraction of these parasites with Triton X-114, contrapsin was found in the aqueous phase of the detergent, suggesting it is unlikely to be an integral membrane protein. Treatment of adult worms with deoxycholate resulted in a change in the electrophoretic behaviour of worm-derived contrapsin. Parallel studies with trypsin suggested this was due to interaction of the serpin with a protease. Using porcine pancreatic trypsin as a model for a putative schistosome protease reacting with contrapsin, we have shown that trypsin, following complex formation with contrapsin, loses immunogenicity. Thus, when contrapsin–trypsin complexes were used as immunogen, the resulting antisera contained antibodies to contrapsin and contrapsin–trypsin complexes only, and none to native trypsin. Thus, epitopes characterizing native trypsin were presumably either masked following complex formation with contrapsin, or their processing and presentation to antigen presenting cells was suppressed, so that an antibody response was not mounted against them. These observations encourage speculation that S. mansoni may be elaborating an immune evasion strategy whereby immunologically sensitive proteases are first complexed with host serpins, which would render them immunogenically inert, and then cleared from the circulation by the host's reticulo-endothelial system. In this way the immune system would be unable to ‘see’ sensitive parasite proteases sufficiently to mount a response against the parasite.