A reduced-complexity detection algorithm is proposed, which is applied to iterative receivers for multiple- input multiple-output (MIMO) systems. Unlike the exhaustive search over all the possible transmitted symbol vectors of the optimum maximum a posteriori probability (MAP) detector, the new algorithm evaluates only the symbol vectors that contribute significantly to the soft output of the detector. The algorithm is facilitated by carrying out the breadth- first search on a reconfigurable tree, constructed by computing the symbol reliability of each layer based on zero-forcing criterion and reordering the symbols according to the symbol reliabilities . Simulations are presented and the good performance of the new algorithm over a quasi- static Rayleigh channel even for relatively small list sizes are proved.