A Global Positioning System (GPS) technology development known as high sensitivity GPS (HSGPS) can significantly improve availability in challenging environments such as in urban canyons where standard GPS performance is extremely poor. However, this technology could produce higher measurement noise, multipath and cross-correlation errors resulting in position errors of hundreds of metres in such cases. The use of internal filtering with “heavy” constraints provides better results in some cases but may result in major biases and overshooting effects in other cases. This paper develops a portable vehicle navigation system by aiding a standalone HSGPS receiver with self-contained inertial sensors and map-matching. Since traditional GPS error estimation methods are shown to be invalid in urban canyon environments for HSGPS, nontraditional data fusion algorithms are needed for augmenting HSGPS with self-contained sensors (MacGougan et al, 2002). A unique map-matching technique is presented that uses both position and velocity measurements and works with a vehicle dynamics model. The model provides a means to assess the quality of HSGPS measurements and therefore an integration mechanism for inertial sensors. The integrated system was tested in a suburban area as well as in a downtown core with buildings of 10 to 50 stories and mask angles of up to 80 degrees. The system performance under severe GPS signal degradation and multipath conditions is presented.