The U.S. Atomic Energy Commission is presently considering lowering the limits of radioiodine release from nuclear power plants by a factor of 30 000 below the current levels listed in its Federal Regulations. This proposed limit corresponds to an environmental concentration of 3 × 10 −15 µCi/ml of air for 131I,with somewhat similar values for other iodine isotopes. This represents a very small amount of radioactivity, corresponding to ∼1 dis/sec in a cube of air 95 ft on a side. We have carried out experiments to determine the lowest radioiodine concentrations that can be measured using high volume collection of air in triethyldiamine-impregnated charcoal and by counting the adsorbed radioactivity with a shielded Ge(Li) counter. For this purpose, we have measured the concentrations of 131I and 133I in air at the fence line of a large boiling-water-reactor plant and obtained 7.6 × 10 −14 and 2.9 × 10−14 µCi/ ml, respectively. In addition, concentration of 131I in the form of fallout at Pittsburgh, Pennsylvania, two weeks after the Chinese nuclear bomb test of June 27, 1973, was found to be 4.6 × 10−15 µCi/ml. The main limitation in the ultimate sensitivity of measurements of this sort lies in the statistical separation of the signal from the background. From our results, we have estimated that 131I and 133 I peaks at the proposed U.S. Atomic Energy Commission limit can be separated from the background with standard deviations of 15 and 30%, respectively, using a total of 7 days for collection and counting. The corresponding figures for a modern state-of-the-art Ge(Li) detector were shown to be 7% for 131I and 15% for 133I. In addition, our measurements possess systematic errors totaling ∼20%.