The purpose of this investigation was to determine the immediate sorption of anhydrous ammonia by soil against diffusion into the atmosphere as influenced by the soil texture, moisture content, pH value, and tilth, when anhydrous ammonia is applied at agriculturally practicable rates directly to the soil. A technique for this determination was devised and evaluated. Approximately 1 kgm of soil was placed in an 800-ml beaker with an injecting tube inserted at the center of the soil to various depths. The beaker and soil were placed within a vacuum desiccator. Anhydrous ammonia was introduced into the injection tube by a special pipette and then the tube stoppered. The desiccator was subjected to evacuation and aspiration cyclicly for 30 minutes (12 cycles), the air being drawn through a standard acid trap which collected the escaping NH3. The excess of acid was back-titrated by standard NaOH solution. The results may be summarized as follows: 1. If the soil is of intermediate texture, moisture content, and pH value, an application of 60 pounds of nitrogen per acre in form of anhydrous ammonia at a depth of 1 to 2 inches is practically all sorbed instantly, and there is little loss by gaseous diffusion. 2. The coarse-textured soil, Plainfield sand, retained the gaseous ammonia with approximately as great efficacy as did the heavier textured soil, Crosby silt loam. This was especially true in the moist, neutral condition. Even in the air-dry condition, this sandy soil contains enough clay (about 6%) to provide adequate sorption capacity for the NH3. 3. The moisture content, air-dry soil as compared to field-moist, exerts only a slight effect on sorption efficacy, and this factor can probably be neglected in field practice. 4. Increase in pH value, effected by the presence of free CaCO3 or the alkalinity due to Na2CO3, did not prevent satisfactory conservation (95 to 100% conserved) of gaseous ammonia by the field-moist soils when the placement was at 2 to 4 inches depth. This was true even at the excessive rate of 600 pounds of N per acre. 5. Coarse tilth did not interfere greatly with the efficacy of sorption in comparisons of aggregates of 2 to 6 mm diameter to fine soil (less than 2 mm). 6. The actual volume of NH3 gas involved is approximately 4.8 liters per lineal foot if traversed by injectors at 15 inch spacing at an application rate of 60 pounds of N per acre. The soil, even when air-dry, sorbed almost instantly over 3 times its own bulk volume (13 times its air-space porosity volume) of NH3, thus illustrating tremendous power of the soil for preventing loss of gaseous NH3 applied as fertilizer. It is therefore apparent that gaseous loss of NH3 from the soil is not an important factor in the use of anhydrous ammonia as a fertilizer.