It is argued that organic agriculture can deliver environmental and ecological benefits, through more efficient use of resources, increased energy efficiency, enhancement of soil quality and improvements in biodiversity and habitat stability. These claims are important in informing the current debate in Europe about the future of land management and future financial support for farming activities. There is however a scarcity of reliable data available on which to base rational decisions. This paper reports the results of a long-term study of two organic rotations in Scotland in which the use of nutrients and other resources has been quantified over a 12-year period. Both rotations operated on a six-year cycle and contained a grass clover ley phase and an arable cropping phase. They differed in that in rotation one the grass clover ley occupied half of the rotational cycle, and in rotation two it occupied two thirds of the cycle. The aim of this comparison between rotations was to evaluate different aspects of environmental ecological and economic performance of contrasting management systems. The yield from the first cereal crop (Avena sitiva L.) is dependent on the soil fertility, which is built up during the grass/clover ley. The grain yield of this crop was between 3.2 and 6.7 t DM/ha. Nutrient cycling in organic farming systems depends critically on the accumulation and turnover of the soil's organic matter, and some studies have suggested that organic management can increase organic matter pool sizes. In our study the total organic matter content of the soil changed very little during the period of measurement, although there was evidence to suggest that organic matter quality varied within rotations. Dissolved organic carbon significantly increased following cultivation of the grass ley. In rotation two there were also changes in particulate organic matter during phases of the rotation with the concentration of particulate organic matter during the cropping phase (19 mg C/kg) being more than double that in the grass ley. Corresponding changes in the stability of soil structure were also observed. Nitrogen inputs to the system were largely through biological nitrogen fixation in the grass ley. Above ground N-fixation varied between 26 and 75 kg/ha over the course of the growing season. Estimated total N-fixation ranged from 70-214 kg/ha/yr. There was a very strong correlation between the dry matter yield of clover and the rate of N fixation, and rates of N-fixation were strongly seasonal and were influenced by both by the month and the age of the grass/clover ley. The availability of nitrogen in the soil was strongly influenced by the rotational management with rotation one containing significantly higher concentrations of nitrate N throughout the study period. We measured nitrous oxide losses from the arable and grassland phases of the rotation between 1996-1999. These showed that losses were consistently less than 1 kg/ha/yr, which is lower than that which would be associated with comparable conventional management systems. Losses of nitrogen by leaching were not measured in the field, but modeled predictions suggest that losses averaged over the 12 years and both rotations were approximately 60 kg/ha/yr. The results of this study demonstrate that an organic farming system can deliver stable crop yields despite relatively low inputs of external resource. Organic matter concentrations were maintained at a relatively high level throughout the study, but the tight cycling of nutrients within the system resulted in relatively low losses. Energy budgeting provides a useful measure of overall resource use in different farming systems. A comparison of energy input/output balances has shown that the amount of energy required to produce a given output in conventional systems is more than double that in this organic rotation. This study highlights the benefit of adopting a systems level long-term appraisal of farming systems and will help in the process of policy development and evaluation in Europe and beyond.