Some dairy cows excrete large amounts of P through their urine; thus, it was speculated that a genetic defect related to their efficiency in uptake of P or recirculation of P could cause such an effect. This speculation was pursued in a cross sectional study on 139 cows (103 Holstein and 36 Jersey) from an experimental herd using repeated sampling of urine (301 samples) to investigate sources of variation in urinary P concentration (Pu). Urine samples were taken on 6 testing sessions spread over 2 mo. Each sample was obtained by mild manual stimulation of the rear udder escutcheon area. The samples were immediately assayed for pH, stored frozen, and assayed for inorganic P and creatinine. Concentrations of P and creatinine in urine, the ratio of Pu to creatinine, and pH were analyzed using a linear mixed model. The model included fixed effects of breed, parity number, and sampling session. Stage of lactation was fitted as Wilmink-type lactation curves. Random effects included additive polygenic ancestry, permanent animal effects, and residual. The distribution of Pu approximated normality except for a single sample with very high Pu and very low pH. This sample came from a cow diagnosed independently with ketosis. For the remaining samples, it was shown that Pu has low to moderate heritability (0.12) and is only moderately repeatable (0.21). Based on a small data set, it is tentatively concluded that individual differences between cows exist in their Pu, and individual differences presumably result from genetic differences. However, it remains unclear if cows with genetically lower or higher Pu will perform better on a low-P diet.