Twenty-four growth hormone-deficient children were treated with growth hormone releasing hormone-40 (GHRH) for 6 months or longer. GHRH (1 to 4 micrograms/kg of body weight per dose) was administered subcutaneously every 3 h (n = 10); or every 3 h overnight only (n = 10); or by twice daily injections (n = 4). Twenty-one children had an increase in growth rate during GHRH treatment. The growth velocities (mean +/- SD; cm/yr) before and during treatment were, respectively: every 3 h 3.5 +/- 1.4 versus 10.0 +/- 2.2, p = 0.0001; overnight only 3.4 +/- 1.0 versus 6.2 +/- 2.1, p = 0.008; twice daily injections 3.2 +/- 1.8 versus 7.9 +/- 2.4, p = 0.06. Using these three modes of GHRH administration, different total daily amounts of GHRH were administered. Regression analysis of average daily dose versus growth velocity revealed a correlation coefficient (r) value of 0.57, p = 0.004. Sixteen children received extended treatment for periods varying from 9 to 30 months. Of these, seven children were treated continuously for 9 months with pump overnight only and 5 for 12 months with pump every 3 h. Their growth velocities were sustained at a similar rate as those observed at 6 months. Six children received both twice daily and three hourly treatments consecutively. The growth velocities were similar during both treatments. Eleven children developed circulating antibodies to GHRH during treatment, however, all 11 had accelerated growth rates during GHRH therapy.(ABSTRACT TRUNCATED AT 250 WORDS)