During storage for 3 years in the dry state, the growth-promoting potency (in rats) of a crude pituitary extract or of purified preparations of growth hormone did not show any detectable change. Storage did not impair the diabetogenic potency (in cats) of the purified preparations, but may have impaired that of the crude extract.When various purified preparations of the anterior-pituitary growth hormone (GH) were assayed for diabetogenic activity in cats, and for growth-promoting activity in rats, no significant variations were found in the ratio of the two activities ('D/G ratio '), even if the preparations had been partially inactivated by treatments such as acetylation [Reid, 1952[Reid, , 1953a. It was concluded that the two activities depend primarily on the same substance or group of substances. Some data for the effect of storage on the potency of GH preparations were obtained in the course of the investigation, and have now been collated.EXPERIMENTAL PROCEDURE Biological assays. Diabetogenic potency was estimated in adult cats [Reid, 1951]. Growth-promoting potency was determined in groups of intact female rats, as described in a previous paper [Reid, 1952], in which it was shown that the slope of the log dose-response hne could be regarded as constant for rats of a certain 'hooded' strain (M.R.C.). Recent small-scale determinations of the slope gave a mean value (7-30+ 1-21 s.E.) similar to that previously found. GH preparations. The GH preparations were obtained from ox anterior-pituitary glands, usually by a method based on that of Wilhelmi, Fishman & Russell [1948], The final freeze-drying of each preparation was performed at a pump vacuum of c. 0-1 mm Hg, with freezing traps immersed in acetone-C02, and was continued for 8-16 hr after removal of water was apparently complete. Certain of the dried preparations were stored for about 3 years (Table 1). Near the end of the investiga¬ tion, a batch of purified GH ('LS') was prepared essentially by the method of Li, Evans & Simpson [1945], and stored in ampoules as for the crude extract listed in Table 1. Since this batch of GH was intended for use as a laboratory standard, it may be mentioned that its growth-promoting potency was found to be 71 % (fiducial hmits 42-121 %, P = 0-05) that of the Armour standard '22KR2'. Tests performed in the laboratories of Messrs Armour and Co.,