Background. It might be possible to estimate the metabolic derangement of cardiac muscle by topical nuclear magnetic resonance spectroscopy (MRS) in vivo without killing the animal. Methods and Results. By use of topical 1H- and 31P-MRS focused on the heart of Syrian hamsters with or without cardiomyopathy (CM; BIO 14.6 strain), the chemical constituents were measured in vivo nondestructively and repetitively at several stages of development of CM. A phantom experiment and two-dimensional plot of chain methylenes (CH2) of lipid/water ratio by 1H-MRS versus creatine phosphate (CP)/[beta-P]ATP ratio by 31P-MRS indicated that signal cross talk from the adjacent organs was negligible. Even before the onset of clinical or pathological manifestation of CM (7 weeks after birth), CH2/water ratio by 1H-MRS was lower in the CM group (7.3 +/- 0.7%) than in control (11.8 +/- 2.0%, p less than 0.05), and it decreased further at the hypertrophic stage (17 weeks, 4.1 +/- 0.7%, p less than 0.05) and the congestive stage (27 weeks, 4.3 +/- 0.9%, p less than 0.05). In contrast, the CP/[beta-P]ATP ratio by 31P-MRS started to decrease at the hypertrophic stage (1.90 +/- 0.18 versus 2.52 +/- 0.24, p less than 0.05) and decreased further at the congestive stage to 1.53 +/- 0.18 (p less than 0.01). These in vivo MRS data were confirmed by both biochemical assay and in vitro MRS analysis in heavy water after the animals were killed. Conclusions. A combination of topical 1H-MRS and 31P-MRS in vivo is promising for the noninvasive and sensitive assessment of cardiac muscle metabolism. Comparison of these MRS studies and biochemical analysis suggested not only the modification of water, lipid, CP, or ATP contents but also the reduction of flexibility or fluidity of lipids in cardiomyopathic heart.