According to the macro rational expectation (MRE) hypothesis, only unanticipated macroeconomic policy has impacts on real economic variables, and anticipated policy changes have no real impacts. This study analyses the effects of the anticipated and unanticipated components of fiscal policy on the US farm real GNP by testing the neutrality and rationality propositions of the MRE hypothesis. The test results show that both the rationality and neutrality propositions are rejected. The empirical findings indicate that the anticipated fiscal policy does have significant effects on farm output. Examination of a specific sector sheds light on the nature of the market and helps ascertain the reasons for the non-neutrality.