Quantitation of articular cartilage by magnetic resonance imaging and three-dimensional reconstruction has been validated and is likely to be a useful outcome measure in clinical trials of arthritis. The cost of magnetic resonance imaging is largely dependent on scanning time. The aim of this study was to compare a fast spoiled gradient-echo sequence magnetic resonance imaging scanning protocol, which takes 5 minutes and 44 seconds, with the standard, previously validated spoiled gradient-echo sequence protocol, which takes 11 minutes and 56 seconds, in the measurement of knee cartilage volume. Cartilage volumes calculated from the standard and fast magnetic resonance imaging sequences were similar. The median absolute percentage overestimation or underestimation of the cartilage volume with the fast sequence for 10 normal subjects was 2.5, 4.1, and 3.2% for patellar, femoral, and tibial cartilage, respectively. For eight subjects with osteoarthritis, the percentages were 4.3 and 3.9% for femoral and tibial cartilage, respectively. The fast spoiled gradient-echo sequence protocol had very high intra-rater and inter-rater reproducibility in normal subjects and in those with osteoarthritis. Because the cost of magnetic resonance imaging scanning is largely determined by scan time, the faster scanning protocol means that the cost of estimates of knee cartilage volume can be significantly reduced, thus increasing the feasibility of this methodology.