A complete solution to joint channel tracking and symbol detection for bit-interleaved MIMO-OFDM systems in a time-varying frequency-selective fading environment is proposed in this research. In the training mode, the Rao-Blackwellised particle filter (RBPF) is used to estimate the phase offset (PO) and the carrier frequency offset (CFO). In the data mode, based on a recursive EM procedure in conjunction with soft decoding, we develop an iterative algorithm that performs the minimum mean squared error (MMSE) channel estimation and the maximum a posterior (MAP) probability symbol detection jointly. The soft decoding module is composed of the list sphere decoder (LSD) and the soft-in-soft-out (SISO) channel decoder based on the BCJR algorithm. The performance of the proposed algorithm is evaluated via simulation. It is demonstrated that the new algorithm has robust performance in the presence of phase noise and the carrier frequency offset.