The Long-Term Research on Agricultural Systems project (LTRAS; www.ltras.ucdavis.edu) at the University of California, Davis was designed to detect and estimate changes in crop productivity, soil quality and environmental trends due to the effects of increasing irrigation, fertilization and cropping intensity/diversity. LTRAS is a long-term research project that addresses the relationships between sustainability and external inputs in Mediterranean and semi-arid regions where irrigated agriculture is characterized by intensive, high yielding crop production. The experiment includes 10 cropping systems which differ in crops (wheat, tomato, maize, and winter legume cover crops, WLCC), amounts of irrigation water, nutrients (particularly nitrogen and carbon) and the form of N/C application: as WLCC, fertilizer, or composted poultry manure. Crop yields and total biomass are measured annually and analyzed for total N and C. Sample archives include time zero soil samples and others collected periodically and yearly plant samples from all cropping systems. Weather data are collected. Irrigation amounts applied are measured using flow meters. Systems rather than single inputs are compared, so each system is optimized individually. One system follows organic farming guide lines and several are biologically-based, relying on nitrogen-fixing legume cover crops for N fertility. The lowest input systems are traditional rain fed wheat/fallow systems without fertilization, while the highest input system is an organic, manure fertility based maize/tomato system irrigated as needed. Sustainability is determined from long-term trends in yield, efficiency in the use of limited resources (such as water or energy), and environmental impact, such as leaching of nitrate and trends in key soil properties such as organic matter. Wheat yields from all three irrigation /fertilization systems have been stable over the last twelve years. Irrigated crops produced more on average, but there was no interaction between irrigation and fertilization. Average grain yields on a dry weight basis for conventional maize were 11990 kg ha-1 yr-1, while for organic they were 7250 kg ha-1 yr-1. Lower yields result in part from different planting dates used between the conventional and organic system. Results for tomato fruit yields were opposite those for grain yields. Conventional tomato yields (fresh weight of harvested fruit) were 56 Mg ha-1 yr-1 while organic yields were 63.3 Mg ha-1 yr-1. Yields are stable for conventional maize but are declining in the WLCC and organic systems. Tomato yields are consistent for all systems. The organic and WLCC maize/tomato system consistently have higher available soil N (NO3, NH4) in both maize and tomato entry points compared to the conventional system. However, well-timed supplemental N fertilizer application in the WLCC produced comparable maize yields to the conventional one, suggesting that N availability was not synchronized to maize needs in the WLCC. Similarly in the organic system, withholding manure input had no effect on yield confirming N availability limited maize yield. Maize/tomato systems achieve comparable yield potential in tomato but not for maize. As a result, N input losses in the 0 to 30 cm soil were lowest in the conventional system where maize yield was highest. After 10 years, soil organic matter levels differed significantly between the conventional and organic systems. Over the first 10 years organic plots had 90 Mg C ha-1 added from crop residues, legume cover crops, and composted manure, while conventional systems received 52 Mg C ha-1, entirely from crop residues. Soil organic matter levels increased in the organic system, but remained stable in the conventional. Across all 10 cropping systems, we found an average residue-C conversion to soil organic C rate of 7.6%. An annual input of 3.1 Mg C ha-1 yr-1 was necessary to maintain initial soil organic C levels. Based on the analysis of soil cores collected in year 10 to 3 m depth, there was no measurable loss of inorganic N (NO3, NH4) in the intensively fertilized conventional or organic corn/tomato systems after ten years of farming. Criteria for sustainability are met variably by differing systems. No cropping system is optimal in all respects.