The present study compared clearance (K) for urea (U), uric acid (Ua), and BSP in relatively equal surface area dialyzers, EX-29 to dual EX-21 in parallel (DEX-21) at Q<sub>B</sub> and Q<sub>D</sub> 300 ml/min, and hollow fiber kidney model 5 (HFK-5) to dual HFK model 4 in series (DHFK-4) at Q<sub>B</sub> 200, Q<sub>D</sub> 500 ml/min. K<sub>U</sub> was 158 ± 20 ml/min, 129 ± 22; K<sub>Ua</sub> was 106 ± 10,96 ± 5, and K<sub>BSP</sub> was 80 ± 11 and 67 ± 12 (p < 0.01) for EX-29 and DEX-21, respectively. K<sub>U</sub> was 173 ± 8 ml/min, 159 ± 6; K<sub>Ua</sub> was 131 ± 6, 117 ± 8, and K<sub>BSP</sub> was 65 ± 9 and 56 ± 7 (p < 0.01) for HFK-5 and DHFK-4, respectively. Comparable changes in outlet pressure had a varied effect on ultrafiltration and solute transport. 3-hour hemodialysis in the same patients showed that EX-29 was better tolerated than DEX-21. The data suggest that dialyzer designs with the same membrane of equal surface area may significantly affect K for both small and middle molecular weight solutes. Comparable changes in outlet pressure may have a varied effect on K in different designs. Due to these variations, comparison of dialysis schedules with different dialyzers on the basis of the square meter hour dialysis to prove validity of middle molecule hypothesis should be interpreted cautiously.