This study has examined the determinants of farmland values in Canada. The empirical results for the period 1959-2004 show that farmland values seem to be disconnected from adjusted earnings per acre regardless of model specification. Differences in model specification can change the interpretation of the importance of government payments in influencing farm land values. If a time trend is included in the land value function government payments appear to have no effect on land values; when the time trend is removed they have a statistically significant positive effect on land values. With respect to the other explanatory variables, the higher the population density, the higher farmland values, indicating that urbanization increases farmland values. Furthermore, increases in real interest rates lower farmland value as the capitalization formula suggests.