The mean North American and world climates have warmed significantly since the beginning of climatologically significant anthropogenic emission of greenhouse gases in the 19th Century. It has been suggested that warming may increase the frequency or severity of droughts. We define and study the statistics of an aridity index that describes the precipitation forcing function of a drought, considering drought to be a season with low enough precipitation to be significant for agriculture. Our aridity index is a reciprocal function of the seasonal precipitation, which is more significant for agriculture than mean precipitation. Using NOAA data from sites in 13 diverse climate regimes in the 48 contiguous United States with time series running over the period 1940--1999 but including two data series from 1900 or 1910, and computing their decadal averages, we search for linear trends in their aridity indices. We find no linear trends significant at the \(2\sigma\) level. At five sites \(3\sigma\) upper bounds on any systematic trends are in the range 1.0--2.8%/decade, while at two sites \(3\sigma\) lower bounds are -0.5%/decade and -2.2%/decade; at other sites the bounds are less restrictive.