The goal of this research was to adapt spatial regression methods to on-farm trials in a farm management context. Different experimental designs and statistical analysis methods are tested with site-specific data under a range of spatial autocorrelation levels using Monte Carlo simulation techniques. Simulations indicated that data usable for farm management decision making could be gathered from limited replication experimental designs if that data were analyzed with the appropriate spatial statistical model.