Urine contains a factor that retards the photo-chemical destruction of riboflavin as revealed by fluorescence measurements. Fluorometric estimation of urinary riboflavin by irradiation technique suffers from an error. Due to this factor inherent in the procedure, and yields results with errors as high as 30% compared to theoretical values. Errors can be minimised to a great extent if results are calculated by including in the analysis two internal standards slightly differing in riboflavin content.