The Chase-Pyndiah algorithm is a soft decoding algorithm for turbo product codes (TPCs) with promising error correction performance. In this letter, we propose an adaptive Chase-Pyndiah algorithm to improve the error correction performance for TPCs with negligible complexity increase. The proposed algorithm adaptively adjusts the weighting factor and reliability factor according to the analog weight of the decision codeword for a component code. Simulation results show that the proposed algorithm can achieve a signal-to-noise ratio (SNR) gain of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.16\sim 0.18$ </tex-math></inline-formula> dB at the bit-error-rate (BER) of 10−6 over additive white Gaussian noise (AWGN) channels.