In this paper, pilot-assisted adaptive prediction iterative channel estimation is proposed for the antenna diversity reception of orthogonal frequency division multiplexing (OFDM) signals. Adaptive prediction filtering is incorporated into iterative channel estimation process with decision feedback and reverse modulation. The filter tap weights are adaptively updated according to changes in the multipath propagation environment. For the adaptation of tap weights, the normalized least mean square (NLMS) algorithm is applied that uses the noisy instantaneous channel gains obtained by reverse modulation and from pilot symbols as the reference signal. The average bit error rate (BER) performance in a frequency selective fading channel is evaluated by computer simulation for the frequency and time multiplexed pilot cases. It is confirmed that adaptive prediction iterative channel estimation provides better BER performance than using fixed-tap filter.