A specific antiserum for human IGF-II has been produced by immunizing rabbits against the synthetic peptide IGF-II(33-40). With this antiserum and IGF-II as tracer a radioimmunoassay for IGF-II has been developed. Cross-reactivity with IGF-I was 0.05% and half-maximal displacement occurred at 2.5 micrograms IGF-II per 1. It was demonstrated that residual IGF-binding protein (IGF-BP) in acid-ethanol extracts interferes with IGF-II measurements and may produce erroneously high values. This interference could be completely blocked by excess IGF-I (25 ng per tube). Utilizing this method IGF-II was measured in subjects at various developmental stages. In newborns, the mean serum level was 237 micrograms/l (N = 56) with a range of 132-430 micrograms/l (5- and 95-percentile, respectively). During the first year of life a considerable increase occurred. Thereafter IGF-II increased only slightly with age from 520 micrograms/l (range 368-735) to 647 micrograms/l (range 507-823) in adults. In patients with growth hormone deficiency (N = 57) IGF-II levels were significantly (P less than 0.001) lower than in controls (mean 261 micrograms/l, range 126-542). It is concluded 1) that residual IGF-binding proteins in acid-ethanol extracts may cause considerable error in IGF-II measurements, and 2) that this interference can be completely blocked by excess IGF-I, if highly specific antisera are used.