A method for monitoring dietary compliance would be useful in treating patients with chronic renal failure (CRF). Nineteen nitrogen balances were measured while patients were eating unrestricted diets containing 6.4 to 15.1 g of nitrogen per day and 14 while patients consumed a 20 to 25 g mixed-quality protein ketoacid-supplemented diet containing 5.2 to 6.6 g of nitrogen per day. Urea nitrogen appearance (U) calculated as the sum of urinary urea nitrogen plus the variation in the body urea pool (using changes in serum urea nitrogen and either the 14C urea space or 60% body weight) was correlated with nitrogen intake (r = 0.84). Both methods gave indistinguishable values for U. Total non-urea nitrogen excretion (NUN) and its components did not correlate with dietary nitrogen. NUN averaged 31.3 +/- 2.1 mg N/kg/day and was not different between the two groups or in patients in neutral compared to those in mildly negative or positive nitrogen balance. Nitrogen balance calculated using estimated U and 31 mg N/kg/day was indistinguishable statistically from measured nitrogen balance. Thus, U varies directly with dietary protein intake and can be estimated using urinary urea nitrogen, SUN, and body weight. Total nitrogen excretion can be estimated accurately as U + 31 mg N/kg/day. From the estimated total nitrogen excretion, dietary compliance of CRF patients in approximately neutral nitrogen balance could be assessed. Furthermore, if nitrogen intake were known, nitrogen balance could be estimated.