Recent research has provided mixed results regarding the presence of a time-varying risk premium in agricultural futures markets. In this paper we test for the presence of a time-varying risk premium and market efficiency focusing on the properties of the underlying data. Specifically, we examine the same markets and period used by McKenzie and Holt (2002) and extend the analysis through 2004. Our results show that accounting for the structural break in the early seventies plays a key role in the findings. In contrast to McKenzie and Holt, we find no evidence of time-varying risk premium in the four commodities analyzed. The corn market appears to be (weak form) efficient. Hogs, live cattle, and soybean meal futures contracts show evidence of inefficiency, which suggests an inability of these markets to incorporate all available information in the futures prices. Our results identify the importance of careful examination of the data as failure to do so can lead to inappropriate conclusions.