Earlier studies usually indicate that farmland prices and cash rents are not cointegrated, a finding that seems at odds with the implications of the present value model. The main objective of this study is to explore whether this absence of empirical support for the present value model can be attributed to the restrictiveness of conventional time series methods. I suggest a panel unit root model with two regimes in which the adjustment process may be characterized by the presence of thresholds and discontinuities reflecting the presence of transactions costs and other barriers to adjustment. Using farmland value and cash rents data for 10 agricultural states of the U.S. between 1960 and 2008, empirical findings give modest improvement over the linear unit root process. It is suggested that there might be a bias caused by cross sectional dependence and an inadequate time span of the data.