Mature articular cartilage is avascular and relies on diffusion of metabolites in order to maintain cellular metabolism and viability. As would be expected for dense connective tissue, the rate of diffusion within cartilage is modest, so that limited delivery of nutrients and removal of metabolic byproducts may play a substantial role in its highly restricted self-repair capacity. Although data are limited, there is evidence that chondrocytes rely heavily on glycolysis, rather than aerobic metabolism, for their bioenergetic demands. Chondrocyte viability in native tissue, and in culture is modulated by oxygen and glucose, as is matrix production itself. The tricarboxylic acid cycle is thought to contribute minimally to the direct energy demands of chondrocytes, but oxygen tension exerts a marked influence on glucose consumption. Lee et al. report a reduction in glycolysis of 40% in anoxic conditions based on lactate production and 25% based on glucose uptake. Lactate, the main by-product of chondrocyte energy metabolism, is an indirect measure of metabolic activity in cartilage. In addition, if production of lactate is greater than diffusive loss, lactate accumulation and the attendant lowering of extracellular pH limits matrix production and cell viability. The relationship of cell viability and matrix production to metabolic state, including substrate availability and byproduct removal, therefore places significant constraints on approaches to cartilage culture and tissue engineering. However, monitoring the levels of glucose, lactate, and oxygen in cartilage is difficult. Indirect, model-dependent, measures of metabolite diffusion rates and nutrient utilization have been performed. Moreover, because indirect measurements of metabolites are made via sampling bath concentrations, concentration profiles within the culture are not available. Therefore, results depend upon assumptions of metabolite mass balance and diffusion characteristics. Direct measurements of metabolites have also been performed, using microdialysis probes positioned at fixed positions within a chondrocyte-seeded gel. Probe placement permitted collection of metabolites as a function of depth into the culture. However, spatial resolution is limited in this procedure, sampling volumes are large and not well-defined, and the arrangement of measurement points is inflexible and determined by pre-defined probe placement. Finally this technique is not suitable for generating metabolite maps within tissue, as that would require many sampling sites. Thus, measurement of glucose, lactate, and oxygen within chondrocyte cultures or within cartilage remains an outstanding problem. Such measurements would permit study of the relationships among tissue viability, mechanical properties, metabolism, and matrix production. It would be of particular interest to characterize the metabolic response to alterations in the glucose and oxygen substrate levels in dynamic experiments. Finally, we note that while measurement of endogenous metabolite levels permits direct metabolic analyses, measurement of the transport of exogenous metabolites within and through cartilage permits assessment of their diffusion characteristics. It is also likely that clinical tissue engineering approaches to cartilage disease will use scaffold-based implants, with the choice of scaffold material having a significant impact on the production of matrix. Xu et al. report increased deposition of GAGs in alginate as compared to agarose in perfusion bioreactors, and increased deposition of collagen under perfused rather than static conditions. This indicates that the mobility of solutes within a construct plays a significant role in the properties of the developing tissue. Localized metabolite measurements would allow us to define relationships between metabolite diffusivity and tissue characteristics; much previous work in this area has relied upon destructive tests. As is the case with metabolites, measurement of oxygen concentrations in tissue is highly problematic. Classical Clark-type microelectrodes must be inserted directly into the tissue, offer limited spatial resolution (ca. 1 mm), consume oxygen during measurements and are sensitive to local flow and diffusion in the liquid phase of the tissue. Fiber optic sensors (optodes) measure dissolved oxygen through its quenching effect on the fluorescence of a fluorophore immobilized on the tip of the fiber. Optodes do not consume oxygen, are relatively insensitive to flow and diffusion effects and offer spatial resolution as high as 50 microns but must also physically penetrate the sample. Finally, as in the case of metabolite measurements with microdialysis probes, electrode-based techniques are not practical for constructing oxygen concentration maps. Previous work in the NMR Unit has led to the development and application of an EPR-based method for oxygen mapping in cartilage. While effective, this technique required availability of highly specialized EPR imaging apparatus. Therefore, the development of alternative, MR-based, oxygen mapping methods suitable for measurements in developing cartilage remains of interest.