The objective of this study was to investigate whether the rate of change in cartilage pathology could be effectively monitored by magnetic resonance imaging (MRI) as part of a longitudinal investigation of an osteoarthritis model in vivo, and to define the minimum requirements necessary to establish disease progression. Magnetic resonance images of the knee of eight male Dunkin-Hartley guinea-pigs were obtained at 8, 12, 18, 24, 30, 36 and 52 weeks of age using a two-dimension spin-echo sequence with a TR of 1500 ms and TE of 40 ms. The total thickness of the femoral and tibial cartilage was measured from those images. Over the same time course, sets of spin-spin relaxation-weighted images were acquired from two additional animals of the same age, from which the T2 relaxation times of water in the articular cartilage were estimated and compared with those of muscle and adipose tissue within the same joint. The cartilage thickness of all the animals increased during the first 6 months, then either stayed thicker (4/8) or became progressively thinner (4/8). Up to 18 weeks of age, the cartilage T2 value was between 23-24 ms but became elevated by 30 weeks and the mean value was more than 40 ms at the end of the study, T2 values for the muscle and adipose tissue remained within the range 30-33 ms, or 47 ms, respectively. We concluded that in this model, cartilage thickness measurements from an magnetic resonance image would not provide a reliable marker to stage osteoarthritis progression partly because the cartilage was so thin in a joint of this size, but also because the changes with time were not linear but biphasic. However, quantitation of the T2 relaxation values may provide a more predictable indicator of cartilage pathology for longitudinal studies because the changes were monotonic and independent of cartilage thickness.