Abstract:The distance sampling is one of the most widely used techniques for estimating abundance and density, particularly for mammals. Managers/researchers sometimes sample unpaved roads assuming that these do not violate important assumptions. However, relatively few studies have attempted to address this, simultaneously comparing behavior between roads and transects, a key issue as behavior can potentially affect detection. Here, we used distance sampling to compare abundance and behavior of agoutis between transects and unpaved roads. We sampled an isolated agouti population during two contrasting seasons. Road densities (347-432 agoutis/ km 2 for dry and wet seasons, respectively) were statistically similar to those from transects (373-322 agoutis/km 2 , respectively). However, road data had a different variance contribution, lower detection probability and poorer model fitting. We also found that agoutis move/interact and forage/feed differently on roads in comparison with transects and were more vigilant in transects than in roads during the first season we sampled (wet). Together, these findings indicated that (1) we detect agoutis differently on these two transect types and (2) road sampling more seriously violates the assumption of randomness of transects in relation to animals. We conclude that while avoiding roads is unfeasible given time and logistical limitations, researchers should be aware of the potential bias, making this clear when discussing and comparing their results.