A new procedure has been developed (2) to study the relation between crop response and response of micro-organisms to soil treatments and to varying fertility levels. The production of CO2 is used as a measure of the response of micro-organisms and the procedure consists in treating soil with mannitol and determining the CO2 at the end of 24 hours. The important feature of this procedure is that only a short time is required to make a test and only one CO2 determination is necessary. In applying the CO2 production method to a large number of soils which had received a great variety of treatments, the following points were brought out: The production of CO2 in soils to which mannitol had been added under controlled laboratory conditions tended to furnish a basis for measuring the nitrogen and phosphorus requirements of soils for cotton. There was a rather high correlation between the response of cotton to lime, phosphorus, and phosphorus and nitrogen and the production of CO2 by mannitol-treated soils. Potassium, however, did not increase CO2 production even where potassium gave increases in crop yields. Micro-organisms (CO2 production) responded to lime and superphosphate only when nitrogen was added. Air drying increased the production of CO2 on rewetting by some soils, but the results were not consistent. On soils which were deficient in CaCO3, the substitution of MgCO3 for part of the added CaCO3 did not consistently increase the production Of CO2. Calcium arsenate had no effect on CO2 production when used in large or small amounts, but intermediate quantities reduced it. Superphosphate did not intensify arsenic toxicity to soil microorganisms nor did FeSO4 alleviate the harmful effects of arsenic as measured by CO2 production. This is contrary to crop response. Soils deficient in copper, manganese, and zinc, as indicated by plant growth, showed no deficiency of these elements by this method. Differences produced in soils by continuous fertilizer and cropping systems were demonstrated with the proposed method. Certain soils were examined by this method which had been studied with the CO2 production method using cellulose (3) as the energy material and a 24-day period. The data obtained with the mannitol method approached more closely the field data and with a great saving in time required to make the tests. Correlation coefficients between crop yields and CO2 production are highly significant. Results by the proposed method tend to correlate with crop response to soil treatment in some cases, but not in others. Further study is required to determine if these inconsistencies may be eliminated by modification of the procedure.