I investigated the effectiveness of activity traps for macroinvertebrate monitoring in shallow, heavily vegetated drainage ditches and explored 2 ways to optimize the use of activity traps for monitoring purposes. I tested the effects of trapping duration (48, 96, and 168 h) and use of attractants (bait and preconditioned leaves). The number of taxa and individuals captured increased with trapping duration. Based on the taxon accumulation curves, deployment times of 48 h and 96 h were equally efficient in capturing new taxa, but a trapping duration of 168 h was much more efficient and resulted in a larger number of taxa collected with every new sample added. Of the attractants offered in the traps, only bait caused differences in the macroinvertebrate assemblage recorded. After 48 h, more predators were captured in traps with bait than in control traps and traps with preconditioned leaves. This effect disappeared with longer trapping duration. Because of their relatively low labor requirements and high level of standardization, activity traps appear to be a valuable tool in lentic biodiversity surveys, especially when deployed for a longer period than has usually been reported. The use of bait is advisable only if capture of specific taxa is required and not for standard monitoring purposes.