It is well established that plasma phosphate (Pp) is largely determined by the renal phosphate threshold, which is best described by the maximal rate of tubular phosphate reabsorption divided by the glomerular filtration rate (Tmp/GFR). For its clinical assessment either direct phosphate loading with simultaneous measurement of GFR is performed, or the nomogram described by Walton and Bijvoet is used. In order to test the validity of the two methods, we compared in 20 infants and 31 children the fasting values of phosphate reabsorption [endogenous phosphate reabsorption/inulin clearance (Tp/Cin) and Tp] with those obtained after phosphate loading [maximal phosphate reabsorption (Tmp) and Tmp/Cin], and both with those derived from the nomogram. In addition the fasting Tp/Cin of 50 infants and 143 children could be compared with the nomogram. The results demonstrate that the directly measured Tp/Cin was the same as the directly measured Tmp/Cin and that the measured Tmp/Cin was correctly estimated by the nomogram. However, the comparison of fasting Tp/Cin with nomogram-derived values showed a systematic error, by which the latter values were higher than those measured. The discrepancy was due to the splay of the phosphate titration curve, which was found by Bijvoet when the ratio of phosphate clearance (Cp) corrected for GFR (Cp/GFR) fell below 0.2. The incorporation of this splay in the nomogram could not be confirmed by data measured in our children. It is concluded that fasting Tp is already "maximal" and that, therefore, no phosphate loading is necessary to estimate Tmp. Furthermore, there is no evidence of a major splay, which makes the nomogram incompatible below a Cp/GFR ratio of 0.2.(ABSTRACT TRUNCATED AT 250 WORDS)