SummaryProperties of human, canine, and porcine heparin and S-35-heparin were first studied. Their electrophoretic mobility through 10 g % polyacrylamide gels, specific activity and their filtration patterns through Sephadex G-200 columns were closely similar. Then, S-35-heparin (0.5 mg) and cold heparin (27 mg) were simultaneously injected intravenously into 5 dogs and their plasma behavior was compared. The plasma half-lives of S-35-heparin averaged 1.18 ± 0. 13 (SD) hr and was identical with that of cold heparin, but the half-lives were much shorter and averaged 0.46 ± 0.08 (SD) hr in 5 dogs when S-35-heparin alone was injected. Studies were next made of heparin affinity to proteins including canine antithrombin III (AT) by the use of Sephadex G-200 chromatography, polyacrylamide gel electrophoresis, and S-35-heparin as a tracer. It was found that S-35-heparin-binding to alpha1, beta, and gammaglobulins and fibrinogen was readily separable upon addition of 5 mg cold heparin, but that the binding to AT was inseparable by addition of cold heparin or by electrophoresis. However, in canine plasma, both S-35-heparin and cold heparin were mostly bound to proteins other than AT, suggesting that this might be the case in vivo. To further substantiate this, studies were made of the comparative behavior of I-125-labeled AT (I-125-AT) and S-35-heparin in dogs with the idea that the plasma half-lives of both should be equal if they were irreversibly bound to each other. The plasma half-lives of I-125-AT averaged 2.10 ± 0.05 (SD) days in 5 male dogs and 1.99 ± 0.04 (SD) days in 5 female dogs, and were much different from the half-life values of S-35-heparin as given above. These results indicate that heparin in vitro is more tightly bound to AT than to other proteins, that heparin in vivo is not irreversibly bound to AT and suggest that it is mostly bound to proteins other than AT in vivo.