4-Nonylphenol (4-NP) is an endocrine disrupting substance (EDS) capable of mimicking the action of 17beta-estradiol (E2). It has been hypothesized that 4-NP in a pesticide formulation is linked to historical declines in Canadian Atlantic salmon (Salmo salar L.) populations, with effects being related to exposure during parr-smolt transformation (PST). To test this hypothesis, Atlantic salmon smolts were exposed to pulse-doses of water-borne 4-NP (20 ug/l), sustained doses of water-borne E2 (100 ng/l) (positive control), or ethanol vehicle (negative control) in mid-May during the final stages of PST. Individually tagged smolts were then sampled at three times (June, July and October) to monitor subsequent growth in sea water and plasma insulin-like growth factor I (IGF-I) concentrations. Smolt weights and plasma IGF-I concentrations were both affected by E2 and 4-NP. The effects of E2 and 4-NP on mean smolt weights were most prominent in July and October [E2 (*98.1 +/- 2.8, *242.3 +/- 10.6 g), 4-NP (*102.1 +/- 3.1, 255.7 +/- 9.5 g), controls (112.5 +/- 2.8, 282.3 +/- 8.8 g)] (P < 0.05), while their effects on mean plasma IGF-I concentrations were most prominent in June and October [E2 (15.0 +/- 1.9, 28.4 +/- 1.8 ng/ml), 4-NP (*14.8 +/- 1.9, *21.6 +/- 1.7 ng/ml), controls (20.0 +/- 1.1, 31.1 +/- 2.0 ng/ml)] (P < 0.05). Additionally, results suggest that the mechanisms of action of E2 and 4-NP involve disruption in the GH/IGF-I axis, and that they may be different from each other. The effects of E2 and 4-NP on growth and plasma IGF-I concentrations observed in this study are ecologically significant because they evoke concerns for successful growth and survival of wild salmon smolts exposed to low levels of estrogenic substances that may occur from current discharges into rivers supporting sea-run salmon stocks.