Abstract: The follow up time is an important parameter for estimating the entry capacity of roundabouts. However, its variability and contributing factors have long been ignored in the literatures. In this study, 171 follow up samples and contributing factors (traffic volume, vehicle position, waiting vehicles behind, vehicle type, and drivers' gender) are collected at a roundabout in Pacific Pines, Australia. It is found that the follow up time is indeed significantly affected by traffic volume, waiting vehicles behind, vehicle type, and drivers' gender. In order to establish the relationship between the follow up time and its contributing factors, an inverse Gaussian regression model is further developed. This relationship could be applied to estimate the entry capacities by taking into account the variability of follow up samples. According to the model, the traffic volume and vehicle types are the most important contributing factors.