Existing traffic networks have to be controlled and regulated efficiently before trying to build new transportation facilities, with an aim to reduce the growing traffic congestion. One of the effective option is signal coordination which helps in improving the uninterrupted flow of traffic. The offset provided for traffic signal coordination is often static assuming constant link travel time making it unfit for varying traffic flows. Present study is aimed at building a dynamic offset model, based on the flow rate and the link length between the intersections meant for coordination. Estimation of travel time required for the vehicles to travel from one intersection to another forms the basis of the offset calculation. The underlying assumption is that the travel time increases with an increase in the rate of vehicular flow, as well as the distance between the intersections. This can be captured by a vehicle detector typically placed at stop line. The relation between travel time and vehicular flow has been verified by conducting studies in congested corridors in Mumbai, India, during peak and off peak hours. An algorithm similar to a vehicle actuated algorithm, but terminate at some common cycle length is used in the development of the models and evaluated using a robust traffic simulator. A non-linear dynamic offset model is hence built, which caters to the real time fluctuating traffic conditions. However, the estimation of vehicular flow rate from the detectors in the field is sometimes difficult and erroneous. To address this limitation, a second model is proposed based on the assumption that the green time provided by the algorithm can be used as a proxy for the detector output and consequently the level of congestion. Accordingly, a dynamic offset model is developed which gives the value of the offset to be implemented based on the green time used and the link length. The models developed are compared with the independent vehicle actuated control and constant offset control. The average delay and travel time are considered as the performance indicator. It is observed that the dynamic offset is successful in improving the coordination efficiency of corridors, even for heterogeneous traffic conditions.