The diagnosis of growth hormone deficiency (GHD) is currently based on clinical, auxological, biochemical and neuro-radiological investigation. Provocative tests of GH secretion using physiological/pharmacological stimuli are required to confirm GHD. The clonidine test (CT) is widely used to assess GH secretory status. In this retrospective study, we analyzed the reliability of CT and the effect of puberty in a large number of children with short stature who had been evaluated for suspected GHD. Data were collected retrospectively from 327 children and adolescents with short stature (204 boys and 123 girls, median age 10.5years (IQR 7.90-12.40) followed in four Italian Paediatric Endocrine Units (Cagliari, Genova, Napoli and Roma) between 2005 and 2013. All children underwent CT as the first GH stimulation test after exclusion of other known cause of their short stature. In 73 prepubertal children and 25 pubertal children, the GH peak after CT was <7μg/L. GHD was confirmed in 87 (37 organic, 50 idiopathic). Six prepubertal and five pubertal patients showed false positive responses. The median BMI-SDS in these children was similar to that of children with GH peak ≥7μg/L, and none were obese. Overall, the prevalence of false-positive responses was 3.3%. The median (IQR) peak GH after CT was similar between prepubertal and pubertal GHD (3.80μg/L [1.7-6.00] vs 3.51μg/L [0.76-5.74]) and non-GHD (13.70μg/L [10.70-18.40] vs 12.40μg/L [9.90-19.25]) children. Our results show that CT is a reliable and safe GH-releasing agent in both prepubertal and pubertal children.