The effects of anticoagulants on the determination of both trypsin inhibitory capacity and the concentration of alpha 1-antitrypsin measured by radial immunodiffusion, and on the alpha 1-antitrypsin phenotype were investigated. These results were compared with those obtained for serum. The following anticoagulants were investigated: sodium citrate; sodium oxalate; buffered citrate; potassium oxalate/sodium fluoride; sodium heparin; and potassium EDTA. It was found that plasmas from all of the anticoagulants, except sodium heparin, resulted in apparently significant decreases of both trypsin inhibitory capacity and concentration of alpha 1-antitrypsin measured by radial immunodiffusion, relative to serum. These decreases were not simply due to dilution by anticoagulants. Using both acid starch gel electrophoresis followed by immunofixation and isoelectric focusing in agarose, no interference was found in the phenotype determination. It is concluded that serum should be used to measure the trypsin inhibitory capacity or the concentration of alpha 1-antitrypsin by radial immunodiffusion, although plasma is also suitable provided that sodium heparin is used as the anticoagulant. The alpha 1-antitrypsin phenotype can be determined with either serum or any of the plasma. None of the anticoagulants employed in this study was present in excess.