We studied healthy volunteers with fat-suppressed three-dimensional (3D) spoiled gradient-recalled acquisition in the steady state (SPGR) to determine parameters that maximize positive contrast between knee articular cartilage and fluid, marrow fat, and muscle; and we compared the technique with conventional MR imaging sequences. The purpose was to determine if fat-suppressed 3D SPGR imaging is useful for detecting abnormalities of the articular cartilages. The knees of 10 healthy volunteers were imaged in the axial plane. Fat-suppressed 3D SPGR imaging was performed with a TR of 60 msec, a TE ranging from 5 to 15 msec, and a flip angle ranging from 20 degrees to 80 degrees. This was followed by a similar set of fat-suppressed two-dimensional (2D) SPGR images, and conventional T1- and T2-weighted spin-echo and multiplanar gradient studies. Contrast-to-noise (C/N) ratios were determined for cartilage versus a saline fluid phantom, marrow fat, and muscle. Optimal parameters were determined both quantitatively and by a blinded subjective analysis. A TE of 5 msec and a flip angle of 40 degrees demonstrated the greatest C/N ratio between the signals for cartilage and for fluid, marrow, and muscle. C/N ratios in the 3D sequences were higher than in the 2D, spin-echo, and gradient series, although the absolute C/N ratio values for the T2-weighted spin-echo sequence were higher than those for the 3D fat-suppressed SPGR sequence. Subjective analysis showed articular cartilage to have a consistent trilaminar appearance, and independent interpreters favored a 3D fat-suppressed SPGR sequence with a TE of 5 msec and a flip angle of 40 degrees. Three subjects with incidental joint fluid had C/N ratios within a 95% confidence range for cartilage versus the fluid phantom. When a fat-suppressed 3D SPGR sequence of 60/5/40 degrees (TR/TE/flip angle) is used, MR images can show high positive contrast between articular hyaline cartilage and adjacent structures. This convenient technique is different from standard MR imaging sequences because it demonstrates greater signal intensity in cartilage than in fluid, marrow fat, and muscle, and because it consistently shows an organized internal architecture of hyaline cartilage. Fat-suppressed 3D SPGR imaging therefore has promise for detecting abnormalities of the articular cartilage.