Recent case-controlled studies have found increases in the serum levels of insulin-like growth factor-I (IGF-I) in subjects who had, or who eventually developed, prostate or premenopausal breast cancers. Since growth hormone (GH) increases IGF-I levels, concern has been raised regarding its potential role as a cancer initiation factor. The epidemiological studies, which indicate an association between serum IGF-I levels and cancer risk, have not established causality. In fact, several alternative explanations for the elevated serum IGF-I levels in cancer patients may be proposed based on human and animal models. First, an effect of IGF-I causing symptomatic benign tissue hyperplasia may result in an ascertainment bias leading to an initiation of procedures resulting in the diagnosis of asymptomatic cancers. Second, elevated serum IGF-I in cancer patients may originate within the tumor (as suggested by some animal studies). Thirdly, serum IGF-I may actually be a surrogate marker of tissue IGF-I levels or of nutritional factors, which are not under GH control and may be involved in cancer initiation. The role of GH in cancer initiation is further negated by the fact that in acromegaly, the incidence of cancer, other than possibly colonic neoplasia does not appear to be significantly increased. Furthermore, GH transgenic mice, with high IGF-I levels, do not develop breast, prostate, or colonic malignancies. It is known that IGFBP-3 can inhibit IGF action on cancer cells in vitro and also can induce apoptosis via an IGF-independent mechanism. Importantly, in addition to increasing IGF-I levels, GH also increases the serum levels of IGFBP-3 and serum IGFBP-3 levels have been shown to be negatively correlated with the risk of cancer in the above mentioned epidemiological studies and in a similar study on colon cancer. These studies suggest that cancer risk is increased in individuals in whom both high IGF-I levels and low IGFBP-3 levels are present. In subjects treated with GH, IGF-I and IGFBP-3 levels both rise together and are not within the elevated cancer-risk range, based on published studies. Long-term studies are needed to assess the potential risks, including the long-term cancer risk associated with GH therapy. These should take into account several factors, including the duration of exposure, the risk magnitude associated with the degree of serum IGF-I elevation, and the adjusted risk based on a concomitant increase in IGFBP-3 levels. Since GH treated patients often have sub-normal IGF-I serum levels, which normalize on therapy, one might predict that their cancer risk on GH therapy should not increase above the normal population. Until further research in the area dictates otherwise, on-going cancer surveillance and routine monitoring of serum IGF-I and IGFBP-3 levels in GH-recipients should be the standard of care. At present, the data that are available do not warrant a change in our current management of approved indications for GH therapy.