Aggregation across time has major implications for interpretation of causality tests. Empirical tests for three corn markets reveal one-way causality to be "fragile" with respect to changes in the level of time aggregation in data. Dynamic multipliers are discussed as one method of checking model specification.