The objective of this work was to evaluate the effects of the insecticides more used in cotton crop in Brazil, on the population of the predator Cycloneda sanguinea (ladybug), and the more appropriate sampling method to evaluate this natural enemy's presence in the cotton. The experiment was installed in field conditions and was conducted with Deltapine Ita-90 cultivar, using the design of randomized blocks, with 15 treatments and 4 replicates. The evaluations seeking to verify the influence of the treatments on the population of Cycloneda sanguinea and the most efficient method for the sampling of this predator in field conditions. Were used the two sampling methods: plant-shaking and Knock down treatment.The results showed that the sampling method for the adults' of the ladybug count, Cycloneda sanguinea, through the plant-shaking was the most efficient, while the sampling method through the knock down treatment it was not shown appropriate for the ladybug adults' monitoring in cotton. In the evaluation of the number of adults of ladybugs in each treatment by the method of the plant-shaking, showed that the insecticides triflumuron, metofenosid, lufenuron and diflubenzuron (IGR), spinosad (naturalyte) and abamectin (avermectin) caused medium percentages of mortality of the ladybug inferior to 28%, while the betacyfluthrin, deltametrina and lambdacyalothrin (pyrethroids), carbosulfan (carbamate), fipronil (fenil pirazol), chrorpirifos (organophosphate) and endosulfan (ciclodien) reached mortality percentages that varied between 46 and 83%. In the final evaluation, accomplished 16 days after the application, the population of ladybugs was reestablished in all the treatments, except for the treatment with the carbosulfan that still differed significantly of the control, provoking a mortality of 40% on the ladybugs.