Four commercially used moisture meters were compared with the official United States Department of Agriculture (USDA) air-oven method to measure the moisture content of 879 samples of corn collected from the 1979 and 1980 harvests. The moisture content of these samples ranged from 11% to 38% (wet basis). All meters were found initially to have statistically significant calibration biases ranging from 1.5% high to 3.5% low. A recalibration before the 1980 harvest reduced both the previous bias and the difference in readings among the meters. Random variations could be traced to three sources: different electrical properties of the grain samples at the same moisture content, meter precision, and oven precision. The proportion of the total variance of each was approximately 85%, 10%, and 5%, respectively. The measured total variance could be estimated as a quadratic function of moisture content, which yielded the minimum value of ± 0.8% within the range of 15.5 - 20.0% moisture content.