In order to improve the performance of intersections in urban areas by offering suitable solutions for the problems found, performance evaluation of intersections is important for design, operations, and planning purposes in traffic management. In order to evaluate the effectiveness of chosen roundabouts and signalized intersections in Bahir Dar, this thesis analyzed capacity performance of intersections using SIDRA INTERSECTION 8.0 program. To identify trends in the performance of intersections in relation to changes in vehicle compositions, the analysis was conducted every 15 minutes over a range of 1-hour consecutive time series. Based on the intersection’s capacity, saturation level, delay, and LOS, and queue length, capacity performance evaluated. Intersections analysis was carried out using an analytical methodology; to determine the LOS, traffic and geometric data were collected for the chosen intersections and analyzed using the SIDRA Software package. The results of the SIDRA software's capacity analysis show that the selected intersections' approaches are providing service closer to and above capacity during peak period traffic flows and have a degree of saturation greater than 0.85, which is the recommended value by HCM. Due to the high values of degree of saturation, delay, and queue length, traffic flow operations at the intersections were unstable. The study's findings indicated that the analyzed intersections have poor performance, and the chosen roundabout and signalized crossings' current conditions need to be improved. The functioning of the intersections in Bahir Dar city had improved as a result of the removal of Bajaj vehicles. When Bajaj’s were removed during rush hour, the average delay in Subatamit approach of New bus station intersection decreased by 228.9second per car and queue vehicle length and distance reduced by 25car and 172.7m respectively, for Ajip intersection average delay decreased by 41.4 sec/veh, for Papyrus roundabout Giorgis approach showed a reduction of average delay of 24min/veh, for Dipo intersection a reduction of average delay of 10min/veh was found. As a result, the performance of signalized and unsignalized intersections is negatively influenced by the traffic mix of three-wheelers.