Highway safety plays a critical role in saving people lives, the increasing number of road accidents and taking millions of people lives all over the world, made highway safety management in first priority for most countries. The main objectives of this study are to analyse High Accident Locations (HAL) by conducting prediction models using (SPSS) program for Traffic accident data, Geometric and Traffic condition data for 30 urban arterial streets in west Amman area. Also, conducting a questionnaire and personal observation to be aware of most of the factors that causes road accidents. Then, conduct level of service of safety (LOSS) to show the rank of safety for each street. Also, Potential Accident Reduction (PAR) criteria was used to identify accident prone locations. Finally, a proposed countermeasures management program was done as a guide for the decision makers on highway safety management. It was found that increasing speed limit will increase the number of fatalities and severe and slight injuries, but will decrease the property damages and the total accidents. Increasing the number of intersections will decrease the number of fatalities and increase the number of slight injuries. The increase in the annual average daily traffic (AADT) will increase the number of severe injuries. Finally, the decrease in street width will increase the number of total accidents and property damages. The questionnaire analysis ranked the causes of road accidents and found that the most effected factors relates to road characteristics. The ranking of the selected streets depending on (LOSS) showed that the level of service of safety is low for all selected streets, and found that 15 streets is considered as HAL according to PAR criteria. The study recommends that international standards should be applied during design and construction of roads. Keywords: Road Safety Management (RSM), Highway Safety Manual (HSM), Level of Service of Safety (LOSS), Safety Performance Function (SPF), Severity Index (SI).