The age dependence of galactose and galactitol excretion was assessed in overnight-fasted galactose-1-phosphate uridyltransferase-deficient patients under dietary treatment (ages 4-34 years; n = 51), obligate heterozygous parents (ages 25-71 years; n = 49) and healthy subjects (ages 3-58 years; n = 215). Urine concentrations were analysed by stable-isotope dilution gas chromatography mass spectrometry. There was considerable interindividual variability. The intraindividual variation, however, was not age-dependent and was rather low. Excretion estimates were calculated from the creatinine-related concentrations using weight-, age- and sex-related creatinine excretion rates. Experimental evidence is presented underscoring the problems inherent in random sampling and substantiating the primary endogenous origin of galactose and galactitol in postabsorptive urine samples. Age-dependent excretion estimates were best fitted to a simple growth-related model assuming an exponential decrease with age until adulthood. According to the model, mean postabsorptive galactose and galactitol excretion in healthy subjects was similar and decreased exponentially from about 1.2 micromol/kg body weight per day in infants to about 0.2 micromol/kg body weight per day in adults. Excretion in heterozygotes was normal. In galactosaemic patients, galactose excretion was in the normal range. Galactitol excretion, however, was enhanced over 50-fold and decreased from a mean estimate of about 64 micromol/kg body weight per day in infants to about 23 micromol/kg body weight per day in adults. The results are discussed with respect to the significance of galactose and galactitol excretion for whole-body galactose removal and with respect to the applicability of urinary galactitol analysis for metabolic monitoring in galactosaemia.