The agricultural mesocosm of Biosphere 2, known as the Intensive Agricultural Biome (IAB), provided food for the inhabitants of the facility during two periods of material closure between 1991 and 1994 (Mission I, September 26, 1991 to September 26, 1993, eight-person crew; Mission II, March 6, 1994 to September 17, 1994, seven-person crew). The design and operation of the mesocosm and preliminary results for food production of the IAB are described for both periods. The overall rate of crop production for the 0.22 ha area (soil depth of 1 m; soil and atmospheric volumes of approximately 2000 m 3 and 38 000 m 3, respectively) sustained both crews. Overall production rates in Biosphere 2 exceeded those characteristic of fertile agricultural land in the most efficient agrarian communities, despite comparatively lower light levels, lack of insect pollinators and unusually dense insect pests. Crop yields were markedly higher for Mission II than for Mission I due, in part, to experience and improvements based on the first closure. The health of the Biospherians is briefly discussed in the context of a low-calorie (1800–2200 kcal day −1 per person for Mission I and 2200–2400 kcal day −1 for Mission II), nutrient-dense diet characteristic of the Biosphere 2 food paradigm. High productivity and biodiversity were due to many factors including high resolution climate control, hyper-intensive agricultural practices, selection and planting of food crops adapted to humid, tropical and sub-tropical conditions, nutrient recycling, intensive pest management, and the superambient levels of atmospheric CO 2 (concentrations up to 4500 ppmv were reported during the 1991 to 1994 occupations). Radiation use efficiency (RUE) for wheat for both periods and a post-Mission II planting were comparable to RUEs observed in other experimental elevated CO 2 settings such as Controlled Ecological Life Support Systems (CELSS) and Free Air CO 2 Enrichment studies (FACE) even though yields were comparatively lower due to low light levels. Integrated management of pests, soil conditions and agricultural practices were key factors in the sustainability of the IAB resulting in minimization of plant loss due to insect herbivory, nematode infestation and reduction in the quality of IAB soils. The use of soils rather than hydroponic systems for the IAB had significant consequences for CO 2, N 2O and O 2 concentrations in the Biosphere 2 atmosphere and rendered primary regeneration technologies ineffective over the periods of closure. The initial high organic carbon content of the IAB soils prescribed by the designers proved to be the largest single source of CO 2 and the largest sink for O 2. The choice of a soil-based compared to a hydroponic-based agricultural system contributed to the accumulation of N 2O to levels as high as 300 times current ambient levels (approximately 310 ppbv). The IAB of Biosphere 2 has the potential, with system improvement, to be a high-yielding, self-sustaining agricultural mesocosm suited for a variety of research endeavors.