Machine-to-Machine (M2M) communication has emerged as a key technology with huge market potential for cellular service providers deploying LTE networks. Addition of enormous number of M2M devices into the cellular networks poses a heavy competition to existing Human-to-Human (H2H) devices for getting radio resources, thereby affecting the performance of the H2H communications. But, one can not treat all M2M flows as low priority and schedule them after H2H flows, as there are many M2M applications like healthcare and tracking which are of high importance and delay-intolerant. Hence, there is a need for class based priority scheduling of the traffic of M2M and H2H sessions in the network. In this paper, we propose a class based dynamic priority scheduling algorithm for uplink transmission of M2M and H2H traffic in LTE. The performance of the algorithm is evaluated by various metrics such as H2H throughput and system throughput and also compared with existing schedulers.