Bioelectrical impedance analysis (BIA) is used to assess skeletal muscle mass, although its application in the elderly has not been fully established. Several BIA modalities are available: single-frequency BIA (SFBIA), multifrequency BIA (MFBIA), and bioelectrical impedance spectroscopy (BIS). The aim of this study was to examine the difference between SFBIA, MFBIA, and BIS for assessment of appendicular skeletal muscle strength in the elderly. A total of 405 elderly (74.2 ± 5.0 yr) individuals were recruited. Grip strength and isometric knee extension strength were measured. Segmental SFBIA, MFBIA, and BIS were measured for the arms and upper legs. Bioelectrical impedance indexes were calculated by squared segment length divided by impedance (L2Z). Impedance at 5 and 50 kHz (Z5 and Z50) was used for SFBIA. Impedance of the intracellular component was calculated from MFBIA (Z250-5) and BIS (RICW). Correlation coefficients between knee extension strength and L2Z5, L2Z50, L 2RICW, and L2Z250-5 of the upper legs were 0.661, 0.705, 0.790, and 0.808, respectively (P 0.001). Correlation coefficients were significantly greater for MFBIA and BIS than SFBIA. Receiver operating characteristic curves showed that L2Z250-5 and L2RICW had significantly larger areas under the curve for the diagnosis of muscle weakness compared with L2Z5 and L2Z50. Very similar results were observed for grip strength. Our findings suggest that MFBIA and BIS are better methods than SFBIA for assessing skeletal muscle strength in the elderly. © 2013 the American Physiological Society.