We present the results of a stated preference study undertaken in Italy in 2017 on individuals’ preferences between an electric car (EC) and a petrol car, with the purpose of assessing the impact of the latent variable EC knowledge on purchasing decisions. We estimate a multinomial, a mixed and two hybrid mixed logit models, with the interaction between EC knowledge, car attributes and additional exogenous covariates. We use three measurement equations to estimate the self-assessed car knowledge, assessed EC knowledge and EC driving experience. We report three main findings. First, the inclusion of EC knowledge improves our capability to explain car choice. Second, the degree of EC knowledge does not change the negative perception respondents have, ceteris paribus, on ECs. Third, the level of EC knowledge influences the importance placed on the attributes of the choice model. Specifically, a higher level of EC knowledge is associated with a lower concern with fast charging station density. Our results are useful for car manufacturers who wish to improve their marketing strategies through tailored advertising efforts, and for policy makers who wish to implement educational campaigns as a means to foster EC uptake.