The majority of energy consumed by contracting muscle can be accounted for by two ATP-dependent processes, cross-bridge cycling and Ca2+ cycling. The energy for Ca2+ cycling is necessary for contraction but is an overhead cost, energy that cannot be converted into mechanical work. Measurement of the energy used for Ca2+ cycling also provides a means of determining the total Ca2+ released from the sarcoplasmic reticulum into the sarcoplasm during a contraction. To make such a measurement requires a method to selectively inhibit cross-bridge cycling without altering Ca2+ cycling. In this review, we provide a critical analysis of the methods used to partition skeletal muscle energy consumption between cross-bridge and non-cross-bridge processes and present a summary of data for a wide range of skeletal muscles. It is striking that the cost of Ca2+ cycling is almost the same, 30-40% of the total cost of isometric contraction, for most muscles studied despite differences in muscle contractile properties, experimental conditions, techniques used to measure energy cost and to partition energy use and in absolute rates of energy use. This fraction increases with temperature for amphibian or fish muscle. Fewer data are available for mammalian muscle but most values are similar to those for amphibian muscle. For mammalian muscles there are no obvious effects of animal size, muscle fibre type or temperature.