Quantitative recoveries of bacteria added to a number of soils were obtained using fluorescein isothiocyanate (FITC) as a stain. Enzyme fluorescence, using fluorescein diacetate or dibutyrate, could not be adapted for the routine counting of metabolizing organisms in soil. Staining with rose bengal gave recoveries of 60–80%.Soil-extract dilution plates revealed 4 × 107 bacteria/g in the 0–10 cm layer in April and 27 × 107 in October. One-third of this number was observed in the 20–30 cm layer. Actinomycete colonies added another 20% to the counts at the 0–30 cm depth but constituted 70% of the total colonies from the 90–120 cm depth. Microscopic counts indicated 2.2–4.6 × 109 organisms/g of surface soil. The number of bacteria decreased in a linear fashion with depth but no relation was found between the numbers obtained by direct microscopy and those by plate counting. However, each of the methods showed a high relationship between the size of the bacterial population present at each depth on the different sampling dates.The observed bacteria, actinomycete spores, and hyphal segments averaged 0.6 × 1 μ. On this basis, the biomass in the top 30 cm, as determined by direct microscopy, ranged from 30 to 76 g/m2 (dry weight basis). This implies, considering the amount of available energy, that the individual cells have enough energy to divide only a few times each year. Direct microscopic counting yields a quantitative picture of the organisms present but cannot separate between inactive cells, spores, and metabolizing organisms. Data from plate counts probably give a truer indication of metabolically active cells.