Thesis (M. Tech. (Civil Engineering)) -- Central University of Technology, free State, 2013 / According to the Road Traffic Management Corporation (RTMC) of South Africa, a large number of accidents involving motor vehicles occur annually on the arterial roads passing through the suburban residential areas of the cities of South Africa. This problem warrants planning and design interventions for the reduction of vehicular accidents and the consequent improvement of road safety on these city roads. Based on this premise, an investigation was conducted to explore the major causes of vehicular accidents, and to develop a set of urban planning and design guidelines to reduce vehicular accidents in suburban arterial roads of a city and to improve the road safety appreciably. The investigation was conducted by considering the suburban areas of Bloemfontein city of Free State, South Africa as the study area. A survey research methodology was followed for this purpose, and data was collected from both primary and secondary sources. Sample surveys were conducted in four different suburban areas of the city to collect primary data and to acquire firsthand information for understanding the scenario at grassroots level. The surveys included household surveys, in order to understand the demographic, socio-economic, and perceptual infrastructural conditions of the study area and their influence on vehicular accidents; road geometrical design parameter surveys; and traffic surveys to understand the road geometry and traffic-related scenarios in the city. In addition, structured statistical data was collected from secondary sources, such as published and unpublished literature and a range of other documents. The data collected was analysed statistically to find the major control parameters influencing vehicular accidents in the suburban arterial roads, and to establish relationships between vehicular accidents and the major control parameters. Based on the analyses, a theoretical linear multiple regression model establishing relationships between the vehicular accidents as the dependent variable and vehicular traffic-related variables (speed of vehicles and average daily traffic), road geometry design variables (road width and median width), and spatial variables (land use and land form in the form of the number of access points from residential areas to arterial routes), was developed to observe the number of accidents under varied simulated scenarios. The simulated model results were employed to develop various policy scenarios to reduce accidents and to improve road safety in the study area. The investigation revealed that, under the composite scenario of the reduction of number of accesses from residential areas to arterial roads, speed, and average daily traffic along with the increase of road width and median width, the occurrence of vehicular accidents in the arterial roads of suburban areas of the city would be reduced and road safety would be improved significantly. It was also observed that residential areas with limited vehicular access from residential areas to arterial roads would have fewer vehicular accidents than residential areas having unrestricted access. Consequently, the number of access points from residential areas to arterial roads in suburban areas of the city would need to be limited, depending on the functions and land use of the area, to improve road safety.