Over an 18-month period serial observations of plasma tyrosine, methionine and urinary tyrosine metabolites were made and compared with urinary succinylacetone excretion in an infant with tyrosinaemia type 1 treated by diet alone. Despite broadly similar profiles there were significant temporal and quantitative differences between each of these metabolic parameters. Only when plasma tyrosine was kept in the low-normal range by strict phenylalanine restriction (10-15 mg phenylalanine/kg body weight) was detectable succinylacetone consistently eliminated from the urine. Urinary succinylacetone is the only measure of metabolite accumulation immediately proximal to the enzyme defect and its routine measurement will allow more effective control of dietary treatment.