An homologous radioimmunoassay (RIA) using the highly purified rat thyrotrophin (TSH) and anti-rat TSH recently made available by NIAMDD is described in detail. Evidence that the assay measures TSH and only TSH includes the following: Treatment of rats with TSH-releasing hormone (TRH) caused a significant increase (averaging 12-fold) and treatment with T4, a significant decrease (averaging 4.5-fold), in plasma TSH. Points for TSH standards and those for dilutions of plasma from TRH-treated rats fell on the same line, and regression lines calculated separately for standards and dilutions of plasma did not depart significantly from parallelism. At 14 days after gonadectomy of male rats, a time when plasma LH and FSH levels are known to be high, the assay showed no increase in plasma TSH. Moreover, reduction of plasma TSH levels by T4 was as great in gonadectomized rats as in controls. Assay of rat LH, rat FSH and rat prolactin, in 7 concentrations each, showed that cross-reaction averaged less than 1 % in all cases. Other workers have calculated values greater than 1 % for TSH contamination of rat LH and FSH. The slopes of regression lines for the 3 hormones tested for cross-reaction did not differ significantly from the slope for TSH standards. This result strengthens the hypothesis that the apparent slight cross-reactions are due to TSH contamination. Findings for T4-treated rats and saline-treated controls show that the homologous RIA is more sensitive than previous, heterologous assays: In previous studies, plasma TSH levels of most or all of rats treated with T4 were not clearly greater than zero. By contrast, in the homologous RIA reported here, values for such rats did not overlap the range of the zero point on the one hand, nor the range for saline-treated controls on the other. Thus, distinct ranges were defined for both normal and low TSH levels. In view of the 1:20 final dilution of plasma during assay, it does not seem likely that non-specific effects of plasma were primarily responsible for the low TSH values of T4-treated rats. Additionally, the useful range of the homologous assay (about 200-fold) is greater than that of earlier assays (about 70-fold, or less). Finally, the use of highly purified rat TSH for standards has avoided inconsistencies previously encountered with bovine TSH standards.