Reducing peak-to-average power ratio (PAPR) is an implementation challenge in orthogonal frequency division multiplexing (OFDM) systems. One way to reduce PAPR is to apply a set of selected partial transmission sequence (PTS) to the transmit signals. However, PTS selection is a highly complex NP-hard problem and the computational complexity is very high when a large number of subcarriers are used in the OFDM system. In this paper, we propose a new heuristic PTS selection method, the modified chaos clonal shuffled frog leaping algorithm (MCCSFLA-PTS). The MCCSFLA-PTS is inspired by natural clonal selection of frog colony and based on chaos theory. Simulation results show that the proposed MCCSFLA-PTS achieves better PAPR reduction than genetic, quantum evolutionary and selective mapping algorithms. Furthermore, the proposed algorithm converges faster than the genetic and quantum evolutionary algorithms.