Insulin-like growth factor I (IGF-I) levels mainly reflect secretion of growth hormone (GH) in the body. The aims of this study were to compare different IGF-I assay methods in healthy individuals, test the reliability of the methods and discuss the utility of IGF-I measurement in adults. The Nichols Institute Diagnostics radioimmunoassay was used to evaluate IGF-I in two random population samples of men and women (aged 25–64 years, n = 392) taken 10 years apart, in 1985 and 1995. This method for IGF-I testing was also compared with an immunoradiometric assay (IRMA) method in 387 men and women participating in the World Health Organization MONICA (MONItoring of trends and determinants for CArdiovascular diseases) Project, Göteborg, Sweden, in 1995. Serum IGF-I decreased with increasing age in both men and women. IGF-I was higher in young women compared with young men in both cohorts, while the opposite was found in the highest age group. Age-adjusted significant correlations were found between IGF-I and smoking, fibrinogen, coffee consumption, lipoprotein (a), osteocalcin and IGF-binding protein 3. The two cohorts showed similar mean IGF-I concentrations irrespective of method. The correlation between the Nichols and the IRMA methods was high: r = 0.93 (p < 0.0001). Based on this and previous studies, population-based IGF-I measurements are robust irrespective of which commercially available method of assay is used. IGF-I levels can be used in diagnosing acromegaly as well as providing target values. IGF-I assay can be used as a complement to stimulation testing in the diagnosis of GH deficiency, and as a tool for GH dose titration.