Plasma and urinary GH responses following acute physical exercise were evaluated in 19 short-statured children (12 males, 7 females, median age: 11.4 yr, age range: 6.1-14.5 yr, Tanner stage I-III, height < or = 3rd centile for age; 7 with familial short stature, FSS; 8 with constitutional growth delay, CGD; 4 with GH deficiency, GHD) and 7 normally growing, age- and sex-matched control children (4 males, 3 females, median age 11.0 yr, range: 7.2-13.1 yr, Tanner stage I-III). All patients and controls underwent a standardized exercise protocol (consisting of jogging up and down a corridor for 15 min, strongly encouraged to produce the maximum possible effort, corresponding to 70-80% of the maximal heart rate) after an overnight fasting. Samples for plasma GH determinations were drawn at 0 time (baseline), at 20 min (5 min after the end of exercise) and at 35 min (after 20 min of rest); urine samples were collected before (0 time) and at 40, 80 and 120 min after exercise. The distance covered by children with GHD during the test was significantly lower (p<0.05) than in the other groups of patients and controls. No differences in the pattern of plasma GH responses after physical exercise were found between children with FSS, CGD and healthy controls, the maximum percent increase (vs baseline) being evident at 20 min (median, FSS: +1125%; CGD: +1271%; controls: +571%). Children with GHD showed a smaller percent increase (+94%) of plasma GH, significantly lower (p<0.01) than those recorded in the other groups. A significant percent increase (p<0.01) of baseline urinary GH following exercise was found in children with FSS (median: +34%), CGD (+18%) and controls (+44%). Children with FSS and CGD showed a gradual increase of urinary GH, reaching the maximum at 80 min, while healthy controls had a more evident and precocious increase (maximum at 40 min). Urinary median GH levels did not change following physical exercise in children with GHD (-5%, not significant). A significant correlation was found between the maximal percent increase (vs baseline) of plasma and urinary GH following physical exercise (r=0.7, p<0.001). In conclusion, our results show that: 1) plasma and urinary GH responses (as well as the distance covered and the number of steps, i.e. the physical performance) to a standardized exercise protocol are similar in children with FSS, CGD and in normal-statured controls, being unable to differentiate among the "normal variants" of growth; 2) children with GHD, unable to accomplish the same performance of the other three groups, show significantly reduced plasma and urinary GH responses following physical exercise. Although the determination of GH responses to pharmacological stimuli remains the definitive tool for the diagnosis of GHD, these preliminary results seem to suggest a potential role of urinary (and plasma) GH response to a standardized exercise protocol as a safe, acceptable first screening test for GH sufficiency also in children, as previously reported in adults.