New peak-to-average power-ratio (PAPR) reduction algorithms for orthogonal frequency-division multiplexing (OFDM) systems are investigated in a probabilistic framework. Specifically, derandomization algorithms based on the Chernoff bound for PAPR reduction are developed by applying the so-called conditional probability method. Our simulations demonstrate that the proposed algorithms outperform several existing algorithms and the computational complexity of the proposed algorithms is found to be significantly less than that of existing algorithms.