In previous work, sustained levels of circulating human growth hormone (hGH) and a highly significant weight increase were observed after electrotransfer of naked plasmid DNA (hGH-DNA) into the muscle of immunodeficient dwarf mice (lit/scid). In the present study, the efficacy of this in vivo gene therapy strategy is compared to daily injections (5 μg/twice a day) of recombinant hGH (r-hGH) protein, as assessed on the basis of several growth parameters. The slopes of the two growth curves were found to be similar (P > 0.05): 0.095 g/mouse/d for protein and 0.094 g/mouse/d for DNA injection. In contrast, the weight increases averaged 35.5% (P < 0.001) and 23.1% (P < 0.01) for protein and DNA administration, respectively, a difference possibly related to the electroporation methodology. The nose-to-tail linear growth increases were 15% and 9.6% for the protein and DNA treatments, respectively, but mouse insulin-like growth factor I (mIGF-I) showed a greater increase over the control with DNA (5- to 7-fold) than with protein (3- to 4-fold) administration. The weight increases of several organs and tissues (kidneys, spleen, liver, heart, quadriceps and gastrocnemius muscles) were 1.3- to 4.6-fold greater for protein than for DNA administration, which gave a generally more proportional growth. Glucose levels were apparently unaffected, suggesting the absence of effects on glucose tolerance. A gene transfer strategy based on a single hGH-DNA administration thus appears to be comparable to repeated hormone injections for promoting growth and may represent a feasible alternative for the treatment of growth hormone deficiency.