Data are presented that demonstrate that ammonia is formed in the course of the analysis of blood samples. The formation has been traced to the decomposition of protein by the action of alkali. The pH established during the diffusion process was of critical importance in the measurements. A technic is described by which ammonia is displaced from blood by a sodium borate buffer at a pH that avoids spontaneous production of ammonia from blood protein. The Seligson-Seligson microdiffusion apparatus is used. Instructions are given for the collection of blood samples and the preparation of reagents and standard solutions.