Tsetse flies are the vector of protozoan parasite of the genus Trypanosoma, the causative agent of human African sleeping sickness and animal trypanosomiasis. Traps such as Nguruman (NGU), biconical and sticky traps are in use for tsetse flies sampling and monitoring. However, there is no evidence regarding their comparative efficiency in catching flies using olfactory cues. Therefore, the present study aimed to evaluate the efficiency of different types of traps in catching tsetse flies at Nech Sar and Maze National Parks, Southwestern Ethiopia. The study was done for six consecutive months from February to July 2019. Briefly, a 3×4 Latin square design was performed, and tsetse flies were collected for three days each month in four different vegetation types, including wood grassland, bush land, forest, and riverine forest. To avoid trapping position bias, rotation of traps has been done every day. Almost all (99.5%) of the flies were Glossina pallidipes and the remaining were G. fuscipes. The latter were present only at Maze national park. NGU traps were the most efficient type with 12.1 flies/trap/day at Nech Sar National Park and it was 2.2 flies/trap/day at Maze National Park followed by biconical and sticky traps. The number of tsetse flies collected by biconical trap was three-fold lower than NGU trap, and it was four-fold lower in sticky trap than NGU trap in both Nech Sar and Maze National Parks. A substantial number (41%) of G. pallidipes were collected from woody grassland (WGL). In conclusion, G. pallidipes monitoring and evaluation activities could consider NGU trap model as it performed better in most vegetation types in the region.