Serum inorganic phosphate was studied in an unselected series of 594 geriatric admissions. Phosphate showed a significant positive correlation with urea and because of the high prevalence of renal impairment the distribution of phosphate was distorted and was log-normal. Removal of cases with urea elevation and other conditions likely to alter phosphate gave a group of 186 patients whose phosphate results were normally distributed with a significantly higher range for females (0.80–1.48 vs. 0.67–1.39 mmole/1). Osteomalacia emerged as the most frequent cause of a low phosphate whilst phosphate elevation was almost entirely due to renal impairment.