Sir.—We read with great interest the article by Lanes et al.1 As the group who suggested to use clonidine as a test for growth hormone secretion in 1979,2 we are very pleased that other groups have found this test useful and simple. Dr Lanes and colleagues propose a standard dose of 50 μg. We have tested a single dose of 25 μg3 and found it sufficient for an initial screening test. If negative (the range of false-negatives being from 20-35), the test should be repeated using a higher dose, preferably 75 μg/sq m, 4 with the dose-response effect of clonidine on growth hormone secretion being linear.3