Activity traps are commonly used to develop abundance indices of aquatic invertebrates and may be deployed with either the funnel parallel to the water surface (horizontal position) or facing down (vertical position). We compared the relative performance of these two positions in terms of numbers of invertebrates captured, species richness of samples, detection rates of specific taxa, and community-level characterizations. Estimates of zooplankton abundance were also compared to quantitative estimates obtained using a water-column sampler. We used a matched-pairs design where 10 pairs of traps (one horizontal, one vertical) were deployed in each of 4 prairie wetlands on 5 dates in 1999. Vertical traps had higher detection rates and captured greater numbers of adult and larval Coleoptera, Hemiptera, Chaoboridae, hydracarina, cladocera, and Copepoda and also produced samples with greater species richness. Horizontal traps captured greater numbers of Amphipoda and Ostracoda and had higher detection rates for these taxa. Estimates of zooplankton abundance with vertical traps also correlated better with quantitative estimates and indicated greater differences between wetlands than horizontal traps. Both traps showed similar relationships among wetlands and changes through time at the community level, but vertical traps were more sensitive to temporal change. Our results indicate that vertical traps outperform horizontal traps and are preferable for obtaining indices of invertebrate abundance.