Objective Dietary salt and fluid restriction is important in controlling fluid balance in CAPD patients. However, it is often difficult to monitor patients' dietary total sodium intake (TSI). Usually, total sodium removal (TSR), the sum of urinary sodium removal (USR) and dialysate sodium removal (DSR), is suggested to represent TSI. In the present study, we investigated the reliability to use TSR as a surrogate to TSI in CAPD patients. Methods Forty clinical stable CAPD patients were closely followed for three months. TSI, USR, DSR and fluid status were measured twice: at baseline and at the end of this study, respectively. Fluid status was evaluated by bioimpedance assessment. Patients with increased sodium intake (group ISI) or decreased sodium intake (group DSI) (both >0.5g/d) were included into this study. Results There were 15 patients in group ISI and 9 patients in group DSI. During the follow-up, although TSI increased in group ISI and decreased in group DSI ( P<0.05), there were no significant changes in USR, DSR and TSR in both groups. No relationship was found between TSI and TSR. Fluid status deteriorated in group ISI and improved in group DSI ( P<0.05). Conclusion: Our study suggests that neither USR nor DSR reflects changes of sodium intake in diet in CAPD patients. Therefore, TSR (the sum of USR and DSR) should not be used to monitor TSI in this patient population.