OBJECTIVE To investigate the effect on the patient's acid-base status of a 39 mmol/L bicarbonate-buffered continuous ambulatory peritoneal dialysis (CAPD) solution. METHODS This was an open, controlled, cross-over, two-center study in 9 patients. After three months of treatment with a 34 mmol/L bicarbonate-buffered solution (t0) patients were switched to a 39 mmol/L bicarbonate-containing solution for four weeks. At the end of the study period (t4) patients were again treated with a 34 mmol/L bicarbonate-buffered CAPD solution for one month (t8). RESULTS Mean venous plasma bicarbonate level significantly increased during the study and decreased at the baseline level during the control period (t0 = 22.94 +/- 2.54, t1 = 26.74 +/- 3.07, t2 = 28.47 +/- 2.68, t3 = 28.11 +/- 3.56, t4 = 28.71 +/- 3.27, t8 = 24.94 +/- 2.56). Arterial blood pH and plasma bicarbonate significantly increased during the study and significantly decreased at the end of the control period (pH: t0 = 7.37 +/- 0.04, t4 = 7.42 +/- 0.04, t8 = 7.37 +/- 0.06. Bicarbonate: t0 = 21.97 +/- 2.57, t4 = 25.85 +/- 2.02, t8 = 21.87 +/- 2.89). The changes in plasma bicarbonate during the study period were inversely correlated with the metabolic acid production calculated from the protein catabolic rate and with the apparent distribution space for bicarbonate (ABS) of patients. CONCLUSIONS The 39 mmol/L bicarbonate-buffered CAPD solution improved the patient's acid-base status. Potential undesirable metabolic alkalosis could be prevented by analyzing the ABS and the metabolic acid production of patients.