A method for measuring 5-L-oxoproline in urine, which involves isolation by short-column chromatography, acid hydrolysis to glutamic acid and enzymic assay of glutamic acid, was used to measure the rate of excretion in normal adults, aged 20 to 45 y. There was no difference in the daily excretion between omnivorous males (217 micromol/d) and females (195 micromol/d). In vegetarian males, urinary 5-L-oxoproline (404 micromol/d) was significantly greater than in vegetarian females (267 micromol/d, P = 0.013). Compared with omnivorous males or females, excretion of 5-L-oxoproline was significantly greater in vegetarian males (P < 0.0001) and females (P= 0.005). When normal adults consumed a diet in which the protein content was controlled at either 4.0 or 6.2 g N/d for 5 d, there was a significant increase in urinary 5-L-oxoproline on d 5, compared with either d 1 or 4. There was a significant inverse linear relationship between the increased urinary 5-L-oxoproline on the fifth dietary day and the nitrogen content of the diet. On the basis of this relationship, when the urinary excretion of 5-L-oxoproline (320 micromol/d) for vegetarians was predicted from an estimate of their dietary intake of nitrogen, the estimate was, on average, close to the measured value (345 micromol/d). As a matter of course, vegetarians excrete more 5-L-oxoproline in urine than do omnivores, and we speculated that this difference might be accounted for by differences in dietary nitrogen and the endogenous capacity for de novo synthesis of glycine.