This paper investigates the driving and charging behaviors of battery electric vehicle (BEV) drivers observed in Shanghai, China. The summary statistics are compared with the observations from the U.S. EV Project. A machine-learning approach, namely self-organizing feature map (SOM), is adopted as a classifier to analyze BEV drivers’ habitual behaviors. The inter-driver heterogeneities are examined in terms of: the distributions of distance traveled per day, the start time of charging, the number of charges per day, distance traveled between consecutive charges, battery state of charge (SOC) before and after charging, and time-of-day electricity demand. It is found that (a) BEV drivers demonstrate conservative charging behaviors, leading to short distances between consecutive charging events; (b) a significant number of BEV drivers in Shanghai charge during daytime; (c) the distributions depicting the driving and charging patterns vary greatly due to the diversity in travel activities among different drivers.