An important aspect of implementing Integrated Pest Management programs is evaluating the compatibility of insecticides with the biological attributes of the natural enemies used for pest control. In this study, we evaluated the effects of selective and non-selective insecticides on the survival and feeding behavior of Hogna cf. bivittata and Lycosa poliostoma (C. L. Koch, 1847), present in soybean crops. We used residual exposure to contaminate adult females of both spider species with the broad-spectrum insecticide thiamethoxam+lambda-cyhalothrin and the selective insecticide methoxyfenozide using concentrations of 20%, 10% and 5% of the maximum recommended field concentration. The survival rate after 120 hours post treatment and the prey acceptance rate in survivors were assessed as ecotoxicological parameters. Methoxyfenozide did not show any disruption to survival of either spider at any of the concentrations evaluated. We did not find significant differences related to mortality between the evaluated concentrations of methoxyfenozide and the control groups for both species. Nevertheless, all the concentrations of broad-spectrum insecticides we used caused mortality rates significantly higher when compared to the selective insecticide. The highest percentages with thiamethoxam+lambda-cyhalothrin were recorded for 20% and 10% concentration. Although we observed moderate mortality rates for the lowest concentrations, the exposed spiders showed a lower acceptance rate when compared to all the concentrations of methoxyfenozide and the control group; the two latter treatments were not significantly different. These results suggest that the selective insecticide can be used in different concentrations without affecting the survival or feeding behavior of either spider species.