This paper proposes a linear transmit filter design to maximize weighted sum capacity (WSC) in multiuser multiple-input–multiple-output (MU-MIMO) systems. The proposed scheme is based on a modified signal-to-leakage-plus-noise ratio (SLNR) criterion, which integrates receiver structures and power allocations into the precoder design and can efficiently exploit unused receiver subspaces. Based on the proposed transmitter (TX) design with receive matched filters (MFs), the WSC maximization problem can be simplified to power-allocation and data-stream selection problems. A power-allocation algorithm for finding a local optimal solution is also proposed and is shown to be obtained by the iteration of closed-form water-filling (WF) solutions. Furthermore, a low-complexity user and substream selection is proposed as an alternative solution to maximize WSC. Simulation results show that the proposed algorithms outperform the conventional scheme and achieve comparable performance to a joint transceiver design, despite requiring simpler receiver structures.