Non-compliance has been reported as a major issue in growth hormone (GH) therapy. We explored the use of urinary GH (uGH) measurements to monitor the GH treatment of 18 children (aged 5-16 years) diagnosed as GH deficient on the basis of history, phenotype, auxology and peak GH concentration during 2 provocation tests of < 15 mU/l. Each child collected 5 consecutive overnight urine samples while on GH replacement schedules, then discontinued treatment for 2 days and collected a further 5 urine samples. The mean mass of uGH excreted on treatment (8.6 ng, range 3.6-13.0 ng) was significantly greater than that off treatment (1.2 ng, range 0.6-2.7 ng; p < 0.01). All uGH values on treatment exceeded the mean nocturnal uGH excretion of normal age- and sex-matched children. The clear distinction between uGH levels on and off GH treatment indicates that uGH measurement would determine whether two or more GH injections had been missed. A series of uGH estimates over a 2-week period may provide a realistic perspective on injection frequency.