Subsurface drip irrigation offers potential for increased water and N fertilizer use efficiency, and decreased groundwater NO3 pollution. Replicated factorial experiments consisting of four rates of N fertilizer application (60-500 log ha(-1)) and three target soil water tensions (SWT) (low, medium, and high) were conducted on subsurface drip-irrigated broccoli (Brassica olearacea L. Italica) during three winter growing seasons in southern Arizona. Objectives were to (i) determine effects and interactions of irrigation water and N inputs on net economic return, residual soil NO3-N, and unaccounted fertilizer N, and (ii) use abstract spatial analysis techniques to simultaneously, evaluate agronomic, economic, and environmental production functions during three growing seasons. Spatial analysis was used to identify overlap of acceptable zones of marketable yield, net return, and unaccounted fertilizer N. Acceptable yields and net return were defined as greater than or equal to95% of maximum predicted response within the range of the treatments, and acceptable unaccounted fertilizer N was defined as less than or equal to40 log ha(-1). During this study, >95% of maximum net return encompassed N rates of 300 to 500 kg ha(-1), and SWTs of 7 to 25 kPa. There was little accumulation of NO3 in the top 0.9 in of soil when less than or equal to350 log N ha(-1) were applied. Unaccounted N increased with excessive N and water inputs, and accounted for as much as 46% of N applied. Overlap of acceptable zones of agronomic, economic, and environmental production criteria was achieved in each year. Areas of overlap were bounded by 300 to 325 kg N ha(-1) and 8.5 to 12 kPa in 1993-1994, 350 to 500 kg N ha(-1) and 11 to 14 kPa in 1994-1995, and 340 to 410 log N ha(-1) and 11 to 24 kPa in 1995-1996.