1. The literature on eosinophil variations under various stresses is reviewed in the light of recent interest in the adrenal cortex as a regulator of blood eosinophil levels. Chamber eosinophil counts are necessary for accuracy in these studies. Phloxine-propylene glycol or eosin-acetone stains with standard white cell pipets and chambers may be employed for simple and satisfactory chamber counts. 2. Statistical formulae are applied to chamber eosinophil counts and marrow differential counts to delineate the theoretic errors. Actual observations differ slightly from the predicted results but substantiate the appreciable intrinsic errors frequently seen. 3. Studies on physiologic variation are reviewed. There are significant minute-to-minute fluctuations as well as a more marked diurnal trend in eosinophil levels with a mid-morning low and a night-time peak. Comparison of single morning eosinophil levels with single noon specimens may be misleading. 4. Oral ephedrine is a potent eosinopenic agent. 5. Repeated tests of the eosinopenic response to ephedrine or epinephrine frequently show significant variation in the same subject. 6. Many patients with various diseases fail to react adequately to ephedrine or epinephrine. 7. Repeated and controlled observations are necessary before conclusions regarding pituitary-adrenal function can be drawn from eosinophil counts alone.