Transit-oriented development has been widely studied in recent years as a means to reduce car trips and promote sustainable transport modes. However, longitudinal studies on the matter are still rare. This paper contributes to longitudinal research of Transit-Oriented Development (TOD) effects on travel behavior by analyzing the evolution of the number of car trips after the implementation of a light-rail metro system in the Porto region (Portugal). As Metro do Porto is a large infrastructure project (a metro network of 67 km), we relied on a macro-analysis performed at the civil parish level. Changes in the number of car trips are evaluated using a difference-in-differences model, extended to a spatial model to account for the metro's spillover effects. These effects became obvious as metro ridership is reported not only in the directly metro-served parishes but also in adjacent non-served parishes. The results highlight the importance of the metro system in reducing the number of car trips, and this effect is visible not only in metro-served parishes but also in the neighboring ones, which are not directly served by the new transport system. Furthermore, we compare the performance of parishes predominantly served with TOD stations to those with transit-adjacent (TAD) and park-and-ride (P&R) stations. We conclude that both station types can reduce the number of car trips, yet only TOD parishes generated significant spillover effects. The importance of other potentially influential factors like building density or socio-economic characteristics is also discussed.