Coloured bowl trapping is a passive sampling method which catches insects through a combination of interception and attraction. The method is commonly used for Aculeata, especially bees. The objective of the present study is to determine a reliable sampling interval for aculeates using bowl traps since an appropriate sampling interval is important to enable the collection of representative samples in Neotropics. For this purpose we installed eight bowl traps in a forest fragment in Southern Brazil, during one year, to compare the sampled diversity in twice and once-a-month sampling interval. We confirmed the general expectation that increasing the sample size results in a proportional greater sampling of the community. Different sampling intervals have a weak impact in rarefaction curves, as these do not reach the asymptote in any cases, and Shannon diversity index was considered statistically equal. Under our results, the twice-a-month interval was considered better than monthly sampling intervals only when we compare the extrapolation data (estimators and extrapolation curve) for Aculeata. The Shannon diversity index and the estimated species number (rarefaction) were considered the same among the treatments, so we cannot suggest a reliable sampling interval from these data.