The objective of this study was to estimate capacity at roundabouts by developing a method to adjust gap acceptance parameters for trucks. Because drivers' gap acceptance behavior is affected not only by trucks in the entry flow but also by trucks in the circulating flow, critical headways were separately estimated for various combinations of vehicle types in the circulating flow at 11 roundabouts in Ontario, Canada; Vermont; and Wisconsin. Because the percentage of trucks was different for different entry legs, the critical headways and follow-up times were estimated at each leg separately. Variations in gap acceptance behavior were also observed at one of the 11 roundabouts for 13 consecutive days to evaluate the statistical significance of differences in behavior between two entry legs. The results showed that a new adjusted critical headway improved the accuracy of capacity estimation and that the critical headways were significantly different between the two legs with different percentages of trucks in the entry flow. The study provided insight into how to capture the effect of trucks on roundabout capacity.