To verify whether combined measurements of GH-dependent parameters might be useful in detecting exogenous recombinant GH (rGH) administration in male athletes from different disciplines.Sixty-six athletes (control group) were sampled for the evaluation of resting IGF-I, N-terminal propeptide of type III procollagen (PIIINP) and telopeptide type I collagen (ICTP). Cut-off values (mean + 2 SD) for IGF-I, PIIINP and ICTP were calculated and arbitrary scores (1·5, 2·0) were assigned to abnormal parameters. By using the sum of individual parameter scores, positive (≥ 3) or negative (< 3) scores were obtained. In addition, a subgroup of six athletes was treated for 3 weeks with rGH (0·09 IU/kg body weight, 6 days/week) and was similarly evaluated at the end of the 1st, 2nd and 3rd week (i.e. 18 samples).Abnormal IGF-I, or PIIINP or ICTP levels were found, respectively, in one, two and four subjects (1·5–6·1%) of the control group (in the younger athletes); only one 19-year-old subject of this group obtained a positive score. Abnormal IGF-I, PIIINP and ICTP levels were found in 61·1–66·7% samples of the treated group. Positive cases were 3/6 at the 1st and 2nd week and 6/6 at the 3rd week. The sensitivity of the screening approach was 50–100% (at the 1st−2nd and 3rd week, respectively) and specificity was 98·5%.This‘first level’ screening test is safe, acceptable and relatively inexpensive. Further additional investigations of‘second level’ (i.e. GH secretory profile, GH response to a GH-releasing peptide) can be retained to validate or exclude rGH administration or for the early diagnosis of infrequent endogenous GH hypersecretion. [ABSTRACT FROM AUTHOR]