While witnessing a remarkable and rapid emergence of Wireless Sensor Network (WSN) that seems to facilitate promising features, congestion problem still remains the key issue that attract very high attention in research community. Congestion not only severely degrades network performance, but also gives adverse quality to various related applications such as real-time monitoring systems that demands strict and stringent requirements. A robust solution is therefore required to mitigate this critical issue, yet to cope with challenging dynamic WSN's nature. Concerning this issue, in this paper, we have investigated a potential solution to alleviate congestion by determining an appropriate packet size in various error conditions. This prospective factor have high potential in improving network efficiency by maximizing the link utilization and minimizing the number of retransmissions, thus ensuring reliability. In addition, the error-prone characteristic and noisy channel in sensor network increase the needs for determining appropriate packet size during data transmission in order to massively reduce congestion and helps providing consistent WSN's performance. This paper investigates the impact of varying packet size in various Bit Error Rate (BER) conditions and highlights other key factors that may help alleviate congestion in WSN. Simulation study demonstrates an overall network efficiency improvement and favourable performance.