The prevalence of clinical signs and symptoms related to fluid retention is high in most studies evaluating the efficacy of GH treat ment in GH-deficient (GHD) adults. This may be a consequence of supraphysiological GH replacement. To examine whether fluid retention is a dose-related phenomenon, we evaluated the impact of various GH substitution doses on body fluid status in 46 GHD men participating in a l-yr, double blind, and placebo-controlled study. The patients were randomized to receive either placebo (n = 13) or GH ina dose of 1(n = 11), 2 (n = 10), or 3 (n = 12) IU/m(2) day, respectively. Treatment was started at one third of the predetermined dose and was subsequently increased by another third every month until the maintenance dose was reached. Tissue hydration was assessed by means of electrical impedance measurements. Normal values were obtained from 128 age- and sex-matched controls. In the untreated GHD state, whole body resistance was abnormally high (observed, 642 +/- 82 Omega; predicted, 550 +/- 31 Omega; P < 0.001). This was mainly caused by an increase in specific resistance of the lean body as a consequence of a reduction in extracellular mater (ECW). The first month of GK treatment was associated with a sharp decline in electrical resistance that was attributed to an increase in ECW.Whole body resistance reached its nadir after 3 months of treatment (517 +/- 72 Omega, i.e. 19.6 +/- 6.5% lower than before treatment; P < 0.001) and did not change significantly thereafter. The GH-induced changes in body resistance were dose dependent. A significant decrease was observed with a dose as low as 0.33 IU/m(2) . day (P < 0.005). However, whole body resistance remained higher than normal in patients receiving less than 0.67 IU/m(2) . day (P = 0.05). Abnormally low resistance values, indicative of overhydration, were observed in patients receiving doses equal to or higher than 2 IU/m(2) . day (P < 0.005). Regression analysis of the pooled data shelved that GH replacement in a dose of 1.10 IU/m(2) . day (95% confidence interval, 0.85-1.45 IU/ m(2) . day) resulted in a normalization of whole body resistance. In conclusion, GH replacement therapy in adults rapidly corrects the preexisting deficit in ECW. This rehydration process is dose dependent and may result in a substantial weight gain. To prevent side-effects related to excessive fluid retention, it is advised to start GH replacement ata dose of 0.5-1 IU/m(2) . day with subsequent monthly increments in steps of O.5-1 IU until a maintenance dose of approx imately 1.10 IU/m(2) . day is reached. This dose is 40-60% lower than that used in most previous studies.