A radioimmunoassay for estradiol-17 beta (E 2 beta) without solvent extraction is described. It can be used for plasma samples with concentrations higher than 10 pg/ml. Tritiated E 2 beta, and a specific antiserum in phosphate buffer were added to plasma samples, the total incubation volume being 0,5 ml. An identical volume of steroid free plasma to that assayed in unknowns (0.050 -0.2 ml) was added to the standard curve. Immunoprecipitation was used to separate bound and free E2 beta and the bound radioactivity counted in the polyproplene assay tube. The calculated regression of E2 beta measured on plasma loaded with excess E2 beta (y = 0.987x / 3.8; R = 0.99) and that of E2 beta measured in the same sample by the direct assay on that of E2 beta found by a reference extraction method (y = 0.998x / 14.9; R = 0.98) as well as the presence of parallelism between the standard curve and different volumes of plasma and acceptable inter and intra assay coefficients of variation show that this method is suitable for the measurement of E2 beta in uteroovarian venous plasma. However, this method cannot be used for peripheral plasma of pregnant animals because it is not specific. The method was found useful in a study on the effect of gonadotrophin pulses on the ovary when many samples had to be analysed. Furthermore, there is a potential for automatization which would facilitate more detailed analyses of ovarian-hypophyseal relationships.