A new adaptive equalizer for wireless systems operating over time-varying and frequency selective multiple-input multiple-output (MIMO) channels is proposed. The equalizer consists of a number of decision feedback equalizer (DFE) stages, each one detecting a single stream. The equalizer filters, as well as the ordering by which the streams are extracted, are updated according to a LS cost function in a VBLAST-like fashion. By taking advantage of the underlying order recursive problem structure, a computationally efficient and numerically robust algorithm is developed based on the Cholesky factorization of the input data autocorrelation matrix. The convergence and tracking capabilities of the equalizer are studied through extensive computer simulations, and its BER performance is evaluated for hostile, time and frequency selective channels.