Four bioassays for use in detecting and measuring insecticide resistance in newly-emerged, unfed adult cat fleas, Ctenocephalides felis (Bouché), were evaluated: horizontal glass, horizontal Nylon 6,6 fabric disk, horizontal cellulose filter paper disk, and vertical cellulose filter paper strip (WHO bioassay). Each bioassay was evaluated using five insecticides: carbaryl, chlorpyrifos, malathion, permethrin, and pyrethrum. LC50s, LC90s, probit line slopes, and slope standard errors were compared. The LC50s on glass were lower than those obtained with the other substrates. This difference was at least an order of magnitude with carbaryl, malathion, permethrin, and pyrethrum. The paper disk and paper strip bioassays produced the highest LC50s and LC90s for fleas treated with carbaryl, malathion, and pyrethrum. With chlorpyrifos and permethrin, the paper strip resulted in the highest LC50s. The nylon disk tended to produce LC50s intermediate between glass and filter paper. On glass, chlorpyrifos generated higher LC50s (2.00 mg[AI]/m2) than permethrin (0.927 mg[AI]/m2) or pyrethrum (0.913 mg[AI]/m2), yet on the paper strip was lower (65 mg[AI]/m2) than permethrin (214 mg[AI]/m2) or pyrethrum (466 mg[AI]/m2). Overall, probit line slopes were highest for glass and WHO. Standard errors of the slope were not significantly different among bioassays. Although nylon disk assay possibly simulates chemical-substrate interactions on carpet, which is a common substrate where cat fleas occur, no single substrate gave acceptably precise probit lines for all chemicals tested, and chemical efficacy depended on the substrate used. Chemical-substrate interactions confound detection of insecticide resistance and chemical efficacy.