A novel resource allocation algorithm for multiuser OFDM systems with proportional fairness is presented in this paper. The algorithm aims to maximum the total transmit rate subject to the overall transmit power constraint and proportional user rate constraint with knowledge of the channel information. The proposed algorithm divides the problem into two steps: bit loading and subcarrier allocation. Bit loading algorithm is easily implemented according to Shannon Theorem with uniformly allocated power on each subcarrier of every user. Then in each iteration of subcarrier allocation, user with least normalized transmit rate has the priority to choose the best subcarrier. The computational complexity of the proposed algorithm is analyzed, as well as the fairness performance which is evaluated by a new proportional fairness index. Compared with two typical suboptimal algorithms, the proposed algorithm can achieve a better balance in proportional fairness, complexity and total transmit rate.