WE have reported¹ that, following the ingestion of radioactive strontium by normal subjects, the specific activity in urine was appreciably greater than that in contemporary plasma. Consequently, we postulated that strontium existed in blood in two different chemical forms. We were unable, however, to find any difference in the specific activity of strontium in plasma ultrafiltrate compared with that of the non-dialysable fraction. This result prompted further experiments which have now revealed an artefact in the original measurements. Heparin, which is known to be a source of adventitious stable strontium², was used in the collection of the blood. An incorrect allowance for this gave us falsely low values for the specific activity of plasma. A complete re-estimation of the relative specific activities of strontium in urine and plasma has therefore been made. The results of these measurements are given in Table 1. The probability of the ratio being greater than unity is 32 per cent, the 95 per cent confidence limits 0.84 and 1.40.