Fostering sustainable transportation is a major objective of many cities. By increasing the use of environmentally-friendly travel modes, city halls can reduce air pollution and street congestion. This necessitates an in-depth understanding of travel mode choice decisions. Identifying and ranking factors influencing mode choices helps city halls better understand and meet the needs and expectations of citizens. Prediction of travel mode is most frequently performed at the level of a single trip. However, the features of an entire tour, i.e. a chain of trips, including the trip of interest, e.g. whether a person visits a school during the next trip of the tour, or even all the travel needs of a person on the day in question may affect mode choices. Hence, in this work, we propose novel tour-related features to complement trip-based features. The tour-related features we propose include not only the features of a tour but also of the first trip of a tour, i.e. the trip starting from home, of the final trip of the tour, and of travel needs of the person during this day. We evaluate the new feature groups with five different trip data sets from London and Warsaw by developing travel mode choice (TMC) models with data including the tour-related features. We use multiple classification methods and recursive feature elimination. Feature importance rankings show that the tour-related features are among the most important features for travel mode choice models. We discuss the most important tour-related features identified based on TMC models.