The effectiveness of selenite cystine (SC) broth, tetrathionate (TT) broth, and Rappaport-Vassiliadis (RV) medium for recovery of Salmonella spp. from 8 highly contaminated foods was determined. RV medium prepared from individual ingredients and incubated at 42 degrees and 43 degrees C was compared with 2 commercial (Difco and Oxoid) media incubated at 42 degrees C. Naturally and artificially contaminated foods were tested under 2 protocols. For Protocol 1, each food was preenriched in the appropriate medium. After incubation, serial 10 fold dilutions of the preenriched foods were inoculated into selective enrichment media and incubated at 35 degrees, 42 degrees, or 43 degrees C. Effectiveness of these conditions was evaluated by most probable number determination of Salmonella spp. recovered. Productivity of selective enrichments did not differ significantly with this protocol, except that with Oxoid RV medium the number of Salmonella cells recovered from most of the foods was significantly reduced. For Protocol 2, twenty 25 g test portions from artificially inoculated foods were examined qualitatively for Salmonella spp. The effectiveness of the broth/temperature combinations was determined by the number of positive tests under each condition. RV medium prepared from individual ingredients and TT broth incubated at 43 degrees C yielded significantly more Salmonella-positive tests from frog legs and lettuce than did SC and TT broths incubated at 35 degrees C or commercial RV medium incubated at 42 degrees C. With pork sausage and ground beef, significantly fewer Salmonella-positive tests were found with Oxoid RV medium incubated at 42 degrees C and SC incubated at 35 degrees C than from other selective enrichments.(ABSTRACT TRUNCATED AT 250 WORDS)