To appreciate the IGF1 sensitivity of breast tumors we detected IGF1-R with a biochemical assay (RRA). We then localized and quantified IGF1-R on frozen tissue sections by histo-autoradiographic analysis (HAA). In some cases, the IGF1 and IGF1-R mRNA expression were studied by Northern blot analysis. We also studied the IGF1 plasma concentration in primary breast cancers compared to controls. IGF1-R (RRA) were found in 87% (n = 297) of the breast cancers. The mean geometric value was 3.87% (specific binding as percentage of total radioactivity); we found a highly significant correlation between IGF1-R and ER on the one hand (P = 0.0001) and PgR on the other (P = 0.0001) (Spearman test). The presence of IGF1-R was associated with a better prognosis, either on relapse-free survival (actuarial analysis: P = 0.004; Cox analysis: P = 0.005) or overall survival (respectively P = 0.003; P = 0.005). The median duration of follow-up was 30 months. By Cox analysis IGF1-R was a better prognostic factor than ER and PgR. In a series of 77 cases of benign breast disease only 47% (36/77) were positive; the mean geometric level was 1.8%. The HAA IGF1-R quantification in 20 breast carcinomas and 12 cases of benign breast disease confirmed the RRA results and demonstrated that the labeling was localized on the epithelial component. In four breast cancers, we did not detect IGF1 mRNA; IGF1-R probe demonstrated two major mRNAs of 11 and 7 kB. Finally we found that IGF1 plasma level was higher in breast cancer patients than in healthy controls of the same age. These results show that IGF1 is implicated in breast cancer growth and suggest that anti-IGF1 treatment might be useful in human breast cancer: for this reason, we and others carried out a phase II clinical trial with somatostatin.