The development of affordable hardware has made it possible to transmit multimedia data over a wireless medium using sensor devices. Deployed sensors span larger geographical areas, generating different kinds of traffic that need to be communicated either in real-time or non-realtime mode to the sink. The tiny sized design of sensor nodes has made them even more attractive in various environments as they can be left unattended for longer periods.Since sensor nodes are equipped with limited resources, newer energy-efficient protocols and architectures are required in order to meet requirements within their limited capabilities when dealing with multimedia data. This is because multimedia applications are characterized by strict quality of service requirements that distinctively differentiate them from other data types during transmission. However, the large volume of data produced by the sensor nodes can easily cause traffic congestion making it difficult to meet these requirements. Congestion has negative impacts on the data transmitted as well as the sensor network at large. Failure to control congestion will affect the quality of multimedia data received at the sink and further shorten the system lifetime.Next generation wireless sensor networks are predicted to deploy a different model where service is allocated to multimedia while bearing congestion in mind. Applying traditional wireless sensor routing algorithms to wireless multimedia sensor networks may lead to high delay and poor visual quality for multimedia applications.In this research, a Priority Rate-Based Routing Protocol (PRRP) that assigns priorities to traffic depending on their service requirements is proposed. PRRP detects congestion by using adaptive random early detection (A-RED) and a priority rate-based adjustment technique to control congestion.We study the performance of our proposed multi-path routing algorithm for real-time traffic when mixed with three non real-time traffic each with a different priority: high, medium or low. Simulation results show that the proposed algorithm performs better when compared to two existing algorithms, PCCP and PBRC-SD, in terms of queueing delay, packet loss and throughput.vi