The productivity of various media used in water analysis for the isolation of the coliform group has been frequently reported in recent literature. Since such studies may not be applicable where milk constitutes a portion of the inoculum, the Most Probable Number was determined in these experiments in a series of liquid and solid media containing milk. The method described by Butterfield (1933) and Hoskins (1933), for determining the comparative productivity of media for coli-aerogenes organisms, was modified, in that 10 instead of 15 tubes of each medium were used and two dilutions were planted instead of three. Sterile whole milk in 1 cc. amounts was added to each tube of liquid media, except standard lactose broth, before the organisms were added. With solid media the number of organisms in the buffer infusion was determined by plating in standard agar. From the results obtained after 24 hours' incubation, two dilutions of the original suspension were plated into each of the trial media in triplicate, such dilutions being used that the plates would show between 30 and 300 colonies per plate.