For the case of imperfect channel estimation, we propose an improved detector for multiple-input-multiple-output (MIMO) systems using an iterative receiver. We consider two iterative detectors based on maximum a posteriori and soft parallel-interference cancellation and propose, for each case, modifications to the MIMO detector by taking into account the channel-estimation errors. Considering training-based channel estimation, we present some numerical results for the case of Rayleigh block fading, which demonstrate an interesting performance improvement, compared with the classically used mismatched detectors. This improvement is obtained without a considerable increase in receiver complexity. It is more advantageous when few pilots are used for channel estimation. Thus, the proposed improved detectors are of particular interest under relatively fast-fading conditions, where it is important to reduce the number of pilots to avoid a significant loss in data rate.