One of the most important elements in highway design is the sight distance available to road users. Sight distance is the visible distance required by a driver to complete a certain maneuver (e.g. coming to a complete stop). If the available sight distance is less than design requirements, the likelihood of a driver safely completing that maneuver decreases. Current methods of measuring sight distance are very difficult, labour intensive and time consuming. Existing methods to obtain sight distance information require field visits or graphic analysis of as-built drawings. This paper proposes an algorithm to automatically extract sight distance from Light Detection and Ranging (LiDAR) data by simulating observer and target points along the virtual highway. LiDAR data is first used to create a surface model of the road. Points representing observers and targets are then created along the highway of interest. ArcGIS software is then used to create lines between the observer-target pairs and obstructions blocking the sightlines are then detected in ArcGIS. A VBA algorithm is written to compute the available sight distance along each sightline. The proposed algorithm was tested on a segment on highway 36 in Alberta. The extracted information was compared to Alberta highway design guidelines and limitations were found to both stopping and passing sight distance on existing highways. It was found that minimum stopping and passing sight distance requirements were not met in two regions. In order to analyse the impacts of sight distance limitations on safety, collision records at limited sight distance locations were assessed. Fixed object collisions and animal collisions were common in those areas, indicating that sight distance limitation could have been a factor in collision occurrence. The method developed in this study could be extremely useful in timely assessment of sight distance on highways, which could, in turn, help address limitations before safety problems arise.