It appears to be widely believed that returns on low quality land are more variable than on high quality land. Using Ricardian rent as the measure of returns and sensitivity to output price as the measure of volatility, we investigate this null hypothesis for three different measures of quality. These are proximity to market, output productivity, and cost efficiency. In all cases, we identify precise conditions on the production technology such that rental volatility varies in a monotone manner with land quality. A method of econometric investigation of the relationship between rental volatility and land quality is developed and applied to Iowa cash rents data collected during 1994-2000. Our preliminary findings provide partial empirical support for the null hypothesis of an inverse relationship between quality and rental volatility with respect to commodity prices.