A computerized urea kinetic model of peritoneal urea transport (PACK-PD) has been developed and used to calculate prescription parameters which would result in the prescribed weekly peritoneal urea clearance (pKpt/V) required to achieve levels of weekly summed renal + peritoneal urea clearance (pKprt/ V) targeted at 1.75 and 2.16. Baseline kinetic data were obtained and analyzed with PACK-PD on 88 patients, and the program then used these data to calculate the required pKpt/V and subsequently the delivered Kpt/V (dKpt/V) from the dialysate collections. A total of 108 prescriptions were written and compared to dKpt/V measured over one to 24 months in the 88 patients. Both continuous ambulatory peritoneal dialysis and automated peritoneal dialysis (APD) were studied (APD consisted of PD+ with one or two diurnal and two to four nocturnal cycler exchanges). The correlation of dKpt/V to pKpt/V showed r = 0.93 with 95% confidence limits (CL) on agreement of +/-20% over a range of pKpt/V 0.52-2.55. The 95% CL on (dKpt/V-pKpt/V) were +/-0.30. We concluded: (1) that the prescription can be modeled as reliably in peritoneal dialysis as in hemodialysis (HD) where dKt/V and pKt/V agree to +/-25%, (2) that any individual weekly dKpt/V may vary as much as 0.3-0.4 from pKpt/V, and (3) that frequent measurement of dKpt/V and adjustment of pKpt/V as needed are required (as in HD) to control mean dKpt/V to within +/-10% of mean pKpt/V.