The efficient market hypothesis, where asset prices follow a random walk and incorporate all relevant information, is often invoked in financial economics. There is some evidence however to suggest that some asset prices do not follow random walks but display long-range dependence. Such systematic behavior of past returns is of interest to traders. This article examines long-range dependence in wheat futures prices using rescaled range analysis and the Hurst exponent. Since this estimate is biased when long-range dependence is absent and its distribution is unknown, a Monte Carlo simulation approach is proposed. Results show that wheat futures prices show no evidence of long-range dependence and there are no profitable trading rules.