Fish plasma/serum contains multiple IGF binding proteins (IGFBPs), although their identity and physiological regulation are poorly understood. In salmon plasma, at least three IGFBPs with molecular masses of 22, 28 and 41 kDa are detected by Western ligand blotting. The 22 kDa IGFBP has recently been identified as a homolog of mammalian IGFBP-1. In the present study, an RIA for salmon IGFBP-1 was established and regulation of IGFBP-1 by food intake and temperature, and changes in IGFBP-1 during smoltification, were examined. Purified IGFBP-1 from serum was used for as a standard, for tracer preparation and for antiserum production. Cross-linking ¹²⁵I-labelled IGFBP-1 with salmon IGF-I eliminated interference by IGFs. The RIA had little cross-reactivity with salmon 28 and 41 kDa IGFBPs (< 0·5%) and measured IGFBP-1 levels as low as 0·1 ng/ml. Fasted fish had significantly higher IGFBP-1 levels than fed fish (21·6 ± 4·6 vs 3·0 ± 2·2 ng/ml). Plasma IGFBP-1 was measured in individually tagged 1-year-old coho salmon reared for 10 weeks under four different feeding regimes as follows: high constant (2% body weight/day), medium constant (1% body weight/day), high variable (2% to 0·5% body weight/day) and medium variable (1% to 0·5% body weight/day). Fish fed with the high ration had lower IGFBP-1 levels than those fed with the medium ration. Circulating IGFBP-1 increased following a drop in feeding ration to 0·5% and returned to the basal levels when feeding ration was increased. Another group of coho salmon were reared for 9 weeks under different water temperatures (11 or 7°C) and feeding rations (1·75, 1 or 0·5% body weight/day). Circulating IGFBP-1 levels were separated by temperature during the first 4 weeks; a combined effect of temperature and feeding ration was seen in week 7; only feeding ration influenced IGFBP-1 level thereafter. These results indicate that IGFBP-1 is responsive to moderate nutritional and temperature changes. There was a clear trend that circulating IGFBP-1 levels were negatively correlated with body weight, condition factor (body weight/body length³ × 100), growth rates and circulating 41 kDa IGFBP levels but not IGF-I levels. During parr–smolt transformation of coho salmon, IGFBP-1 levels showed a transient peak in late April, which was opposite to the changes in condition factor. Together, these findings suggest that salmon IGFBP-1 is inhibitory to IGF action. In addition, IGFBP-1 responds to moderate changes in dietary ration and temperature, and shows a significant negative relationship to condition factor.