The emergence of infrastructure such as Dedicated Short Range Communication (DSRC) allows us to establish ad hoc vehicular networks in which range measurements between the moving vehicles can be quantified and included as part of the integrated positioning solution. This new information source together with Global Navigation Satellite System (GNSS) forms a platform for robust position estimation that can meet the strict performance requirements of a range of road safety systems and services. Cooperative Positioning (CP) is an important application of DSRC that has a crucial role for the reliability of safety-related applications, because it can provide consistent below metre positioning accuracy. CP was analysed with Cramer Rao Lower Bound (CRLB) and proved competent and reliable for the position information in vehicular environments. CP is a positioning solution at no extra cost and with no major implications to DSRC system developers. CP can potentially lead to reliable and consistent 1-metre level positioning accuracy. There are of course, challenges with ranging. Received Signal Strength (RSS) delivers really poor range estimates especially in rough vehicular environment, leaving Time of Arrival (TOA) the only viable option for CP ranging. TOA needs synchronization of DSRC boards that is an ongoing research agenda. In addition to effective ranging, CP requires an efficient positioning algorithm. A modification of non-classic Multi-Dimensional Scaling (MDS) for positioning of vehicles in a cooperative manner was presented. The modifications bear two improvements to the non-classic MDS to suit the vehicular networks: filtering capability for moving nodes and fast convergence, which are both important in the topologically fast changing vehicular environment. Map information fusion and filtering capabilities were introduced. A novel approach to state covariance estimation was presented, which makes possible the filtering of node positions over time and leads to better and smoother position solutions. It was shown via simulation that a special blend of map information and iterative MDS algorithm leads to low computational complexity. The Multi-Dimensional Scaling Filter (MDSF) algorithm with Map-Matching (MM) has almost same computational complexity as an Extended Kalman Filter (EKF) but with better performance. A Netlogo road network simulation platform was used to demonstrate the performance of MDSF. Four different traffic conditions were considered, varying between slow moving (heavy) and fast (sparse). The simulation demonstrated a better performance for MDSF relative to EKF in all traffic conditions of about 20 cm on average. Typical Global Positioning System (GPS) outage patterns were studied and applied in the simulation model. Then the robustness of both MDSF and EKF algorithms, and generally the robustness of CP via CRLB, to GPS outages in the presence of high rise buildings were examined. Both MDSF and EKF proved to gap the GPS outages flawlessly.