The relationship between plasma transforming growth factor-beta 1 (TGF-beta 1) and second primary breast cancer was explored in a prevention trial of the synthetic retinoid fenretinide (N-(4-hydroxyphenyl)retinamide; 4-HPR). Plasma concentrations of TGF-beta 1 were measured by radioimmunoassay in plasma obtained at randomisation and after approximately 1 year of intervention in 28 women treated with 4-HPR and 27 untreated controls with stage I breast cancer. Baseline and 1 year growth factor concentrations were not significantly different between treated and control groups. After a median follow-up of 65 months, an increase in TGF-beta 1 over 1 year was the only significant, independent predictor of a shorter survival free from secondary primary breast cancer. Moreover, the change in TGF-beta 1 levels had a tendency to influence the treatment effect on second breast cancer incidence. Our data suggest that the role of plasma TGF-beta 1 as a surrogate endpoint of breast carcinogenesis should be assessed further.