The three techniques allowing the noninvasive study of cardiac metabolism, namely magnetic resonance spectroscopy (MRS), positron emission tomography (PET) and single photon emission computed tomography (SPET), all use external detection with stable or radioactive isotopes. These techniques yield different information. PET is quantitative and very sensitive, and therefore only tracer amounts of molecules need to be injected. It allows neurotransmitters and receptors to be studied and a global view of metabolism (oxygen consumption, glucose and fatty acid utilization) to be obtained. SPET also has good sensitivity, but uses gamma-emitting isotopes of heteroatoms. Their longer half-lives allow follow-up for hours or days. MRS is based on stable elements with high (hydrogen 1, phosphorus 31, fluorine 19...) or low (carbon 13, Deuterium) natural abundance. It has very low sensitivity and only millimolar concentrations of substrates can be detected, but various parts of metabolism can be studied. The in vivo measurement of myocardial concentration of substances has many problems that are common to all three techniques (measurement of the volume, measurement of the quantity of each molecule, resolution, partial volume effect, improvement of the signal-to-noise ratio, movement of the organ). The complementarity of the techniques is illustrated by their applications to the study of cardiac metabolism. For instance, the energy metabolism can be studied by 31P-MRS, which detects the high-energy compounds ATP and phosphocreatine, and 13C-MRS yields information on the tricarboxylic acid cycle activity. PET and SPET allow the utilization of fatty acids, the normal fuels of the heart, to be studied. During ischaemia, PET with 18F-fluorodeoxyglucose (18FDG) can determine the glucose consumption and 1H-MRS shows the increase in lactic acid, reflecting anaerobic glycolysis. Comparison of the use of acetate labelled with 11C for PET or 13C for MRS shows the potentials and limitations of each technique. Myocardial perfusion can be evaluated directly with various PET tracers or indirectly with thallium 201 or various technetium-99m-labelled tracers by SPET. No MRS marker of perfusion is so far clinically available. Mainly SPET and PET are used clinically for the investigation of ischaemic heart disease as well as cardiomyopathies, but some initial results using 31P-MRS are being obtained.