The introduction of a carsharing service reduces both private car ownership and total vehicle miles traveled. Due to the environmentally friendly modal shifts, policy makers are keen to increase market share of carsharing. While past research has focused on both the users of carsharing services and their general mode choice, little is known about access trips to vehicles of station-based carsharing services. However, because access to a mode is critical for its success, gaining insight into the way people experience access trips is an important step towards increased mode acceptance. In this paper we present model results based on a stated choice survey regarding access trips to carsharing stations conducted by users of a regional carsharing provider in Germany. After giving a brief overview of the conducted survey and the results of a descriptive analysis, we present the results of multiple multinomial and mixed multinomial logit models. Results of the multinomial logit models show that trip-related variables are the most important determinants for access mode choice, while only a few socio-demographic parameters are significant. We estimated mixed multinomial logit models to find out how consistent respondents answered across the choice situations. ... mehrThe results show that respondents were not always consistent across their choices and would, e.g. choose public transportation even if that entailed long distance travel or waiting times. Our findings are consistent with both research regarding public transportation access and the small pool of research regarding access of carsharing vehicles.