The replacement of growth hormone (GH) radioimmunoassays with a variety of more specific immunometric methods in diagnostic service laboratories has led to a worsening of between-laboratory agreement, reflecting differences in method bias. Incorrect calibration and differences in specificity are important causes of method bias, but the impact of this on interpretation is not clear. In order to maximize the diagnostic reliability of GH testing for small stature, manufacturers should carefully calibrate their methods against the appropriate GH International Standard, and should use antibodies of broadly agreed specificity. Laboratories performing GH tests should participate in a reliable External Quality Assessment (EQA) scheme and guidelines for investigation that incorporate normal GH responses should be agreed.