The impact of weather conditions on traffic safety is a topic that has attracted considerable interest in the literature. In this research, an integer autoregressive model (INAR) is used to estimate the effects of weather conditions on four traffic safety categories: vehicle accidents, vehicle fatalities, pedestrian accidents and pedestrian fatalities, using 21 years of daily count data for Athens, Greece. The results suggest that the most consistently significant and influential variable is mean daily precipitation height along with its lagged value. It is found that, contrary to much previous research, increases in rainfall reduce the total number of accidents and fatalities as well as the pedestrian accidents and fatalities, a finding that may be attributed to the safety offset hypothesis resulting from more cautious and less speedy driver behaviour. Similarly, temperature increase was found to lead to increased accidents.