Concern over the status of bees has increased the need to inventory bee communities and, consequently, has increased the need to understand effectiveness of different bee sampling methods. We sampled bees using bowl traps and netting at 25 northwest Indiana sites ranging from open grasslands to forests. Assemblages of bees captured in bowl traps and by netting were very similar, but this similarity was driven by similar relative abundances of commonly captured species. Less common species were often not shared between collection methods (bowls, netting) and only about half of the species were shared between methods. About one-quarter of species were more often captured by one of the two collection methods. Rapid accumulation of species was aided by sampling at temporal and habitat extremes. In particular, collecting samples early and late in the adult flight season and in open and forest habitats was effective in capturing the most species with the fewest samples. The number of samples estimated necessary to achieve a complete inventory using bowls and netting together was high. For example, ≈72% of species estimated capturable in bowls were captured among the 3,159 bees collected in bowls in this study, but ≈30,000–35,000 additional bees would need to be collected to achieve a 100% complete inventory. For bowl trapping, increasing the number of sampling dates or sampling sites was more effective than adding more bowls per sampling date in completing the inventory with the fewest specimens collected.