Abstract Growth hormone (hGH) is a measurand belonging to ISO category 4, indicating intrinsic unavailability of a reference measurement procedure and primary standard material. Large between-method differences have been raising confusion, especially in the interpretation of results of stimulation tests for exclusion of juvenile growth hormone deficiency. Within the framework of the external quality assessment scheme (EQAS) of the SKML (Dutch Foundation for Quality Assessment in Clinical Laboratories), attempts to reduce between-method variation of hGH measurements have been made, starting in 1994 with an inter-laboratory comparison of 9 different immunoassays by using a panel of sera and standard materials available at that time. Methods appeared to differ from each other largely in a systematic, sample-independent manner. These systematic differences are reflected in the hGH measurement results obtained in commutable sera. A commutable serum pool was introduced as a consensus reference material, permitting correction of each method's results to a common scale. Pair wise comparisons (“twin studies”) were carried out to investigate and corroborate the effectiveness of this material for harmonization. A significant reduction of the between-laboratory coefficient (CV) of variation from 22 to 9.0% was attained.