Changes and trends in farm productivity have been of intense interest to many involved with agriculture. This study used data envelopment analysis (DEA) to estimate the output-oriented Malmquist total factor productivity (TFP) index from panel data for 1993-2006 for farms in Southern Minnesota. Bootstrap methods were used to estimate confidence intervals for the productivity, efficiency change and technical change indices. The model included three inputs (labor, land and immediate expenditures) and six outputs (corn, soybean, milk, hog, beef, and nonfarm income). Productivity growth was found to be positive during the period, with an average annual productivity growth of 6.6 percent. However, TFP growth has been slowing down in recent years and indeed negative in 2000/01, 2002/03 and 2005/06. In the second stage of the analysis, the significance of various factors that might affect farm performance was estimated. Farm size (as measured by the log of farm income) was correlated with higher productivity which may help explain the increase in farm size in Minnesota farms in recent years. Government subsidies were found to have a negative impact on farm performance supporting the argument that agricultural subsidies may create disincentives for farmers to improve their productivity and efficiency. A higher nonfarm income ratio was positively related with higher productivity growth. A higher proportion of hired labor has a negative effect implying family labor is more crucial than hired labor in improving productivity.