In Intelligent Transportation Systems (ITS) applications, GPS measurements captured by vehicles are commonly integrated in a geographic information system (GIS) environment to determine vehicle routes. However, the map-matching problem arises when GPS measurements are associated to incorrect road segments on a digital map. Topological map-matching algorithms (TMMAs) have been widely used to successfully solve this problem. This study proposes two pre-processing techniques that are implemented prior to executing a TMMA in an offline context for improving the quality solution and execution times of the algorithm. The first pre-processing technique finds the optimal buffer size for selecting potential road segment for snapping locations of the GPS points. The second pre-processing technique compares the vehicle heading and the direction of the road segments candidates and selects those candidates that are within a heading tolerance from the vehicle heading. Both pre-processing techniques are tested and compared using GPS measurements collected by cargo vehicles in the commune of Renca in Santiago, Chile. Overall, results show that the identification of an optimal buffer size for each vehicle route and the use of a heading tolerance of 20° improve solution quality and computing times. Finally, the pre-processing techniques yield improve results for different sampling intervals of the GPS measurements.