Studies on the relation of cropping to numbers of bacteria in soil were carried on in an experiment designed for statistical control. Plots were cropped as follows: fallow, three; wheat, three; barley, two; sugar beets, two; soybeans, two; and corn, two. The estimate on each plot is based on counts from 24 plates. A series of from 17 to 20 such estimates, made at weekly intervals from each plot, is available for the analysis. Moisture data are available also. The effect of cropping on numbers of bacteria is shown for each plot as the multiple regression of bacteria on days and moisture reduced to two dimensions by the elimination of moisture.Bacteria respond to a growing crop. This response is similar in replicate plots of one treatment; but is markedly different with different crops.