Considerable progress has been made over the past decade in developing imaging methods for quantifying cartilage mor- phology. Over the coming decades, these methods will likely be supplemented by non-destructive "molecular" or "bio- chemical" imaging by MRI to provide an adjunct or surrogate for the destructive histological and biochemical assays used today. In this talk, we will provide a brief overview of emerg- ing strategies for assessing cartilage molecular state – i.e., its biochemical composition and architecture. Importantly, the macromolecular status of cartilage is inextricably linked to the functional integrity of the tissue. In particular, studies suggest that solid/fluid volume fraction, collagen composition, molecular structure and organization, and GAG composition are key macromolecular features that they have interdependent effects on functional integrity of cartilage. Accordingly, these macromolecular features have been the focus of efforts to develop non-destructive imaging methods. Techniques under development range from those specific to a given biochemical species (water, collagen ultra- structure, and GAG) to those that appear through correla- tive studies to have image contrast that is modulated by a particular combination of macromolecules. Our talk will briefly comment on many of the methods that have been used, and then look in some depth at several. Water content can be measured directly and specifically using proton density imaging. The practical challenge lies in the sensitivity, because the variation of water content across disease states is less than 5%. As noted below, there are some MR contrast mechanisms that can overemphasize this significant but small change in proton density – albeit at the cost of specificity. Collagen orientation can be elucidated using microscopic MRI (ÌMRI) by measuring the angular dependence of T2 images. This angle-dependency is referred to as the magic angle effect, and it arises because the preferred orientation of the matrix (primarily collagen) confers a preferred orientation to the water dipoles that are probed by T2 imaging. Although there are many factors that can affect T2, the angle depend- ence of T2 (magic angle imaging) provides specific informa- tion about the macromolecular (collagen) ultrastructure. Methods developed to measure GAG, sodium imaging and the dGEMRIC method take advantage of the fact the GAG is charged - in fact, they can more accurately be said to be assessing fixed charge (the charge "fixed" to the macro- molecule, which in cartilage is essentially all due to GAG). These methods have considerable dynamic range and ensu- ing sensitivity to changes in [GAG] that one might observe in vivo. ([GAG] can range from around 60 mg/ml tissue water in normal native cartilage to less than 10 mg/ml in diseased, repaired, or engineered cartilage.) In addition, the well- developed biophysical understanding underlying the meth- ods permits, in principle, a measurement that can be quanti- tatively related to [GAG]. Thus, charge-based methods for evaluating cartilage GAG are (or can in principle be) sensi- tive, specific, and quantitative. Thus, in overview, proton density, magic angle, and charge-based imaging are each methods that are grounded in and supported by biophysical models that enable a pre- dictable relationship between MR image and water content, collagen ultrastructure, and glycosaminoglycan concentra- tion, respectively. This biophysical basis confers a relatively high specificity. For obvious reasons, specific methods are appealing; how- ever, in the long run it is likely that considerable insights and important clinical and pathophysiological information will be gained from methods that are straightforward and practical, and not necessarily entirely specific. It is also likely that, when combined with information from a high-specificity method, unambiguous conclusions about the biochemical state can be derived from otherwise non-specific methods. We therefore offer a brief mention of many of these approaches. Among the earliest methods to be considered to provide 365 J Musculoskel Neuron Interact 2004; 4(4):365-368 Molecular (and functional) imaging of articular cartilage M.L. Gray 1,2 and D. Burstein 1,3 1 Health Sciences and Technology, Harvard and MIT, 2 New England Baptist Bone and Joint Institute, Boston, MA, 3 Department of Radiology, Beth Israel Deaconess Medical Center, Boston, MA, USA Keywords: Cartilage, Imaging, MRI, Glycosaminoglycan, Arthritis Review Article Hylonome M.L. Gray has no conflict of interest. D. Burstein is a consultant for Pfizer, Inc. Corresponding ·uthor: Martha L Gray, E25-519, 77 Massachusetts Avenue, Cambridge, MA 2139, USA E-mail: mgray@mit.edu Accepted 4 August 2004