Traffic congestion is a major problem around the globe. The prime reason for congestion is unavailability of traffic infrastructure to meet the traffic demand. Road users are forced to face undesirable delay, which influences the economy, environment and health. The present study examines the congestion in the urban roads of Bhubaneswar, a smart city in India. Travel time for various categories of vehicles was estimated and congestion indices in terms of buffer index were evaluated. Multiple linear regression modeling has been used to evaluate the congestion parameters. The p-value for all the independent variables in the developed model is < 0.05. Four elements, namely Strict traffic law implementation, Adequate parking facilities, Decentralization and Controlling the road side activities are required to improve the serviceability and mobility of urban road networks.