<p>We investigated the physical characteristics of older adults with suspected locomotive syndrome (LS). Subjects were 378 community-dwelling older adults (73.6 ± 5.3 years; women, 53.2%). To measure their physical function, we used tests of grip strength, one-leg balance with eyes open, 5-time sit-to-stand, timed up and go, and 5-m habitual walk. We used the "loco-check", a 7-item questionnaire, to measure LS. Two grouping methods were used. In comparing non-LS (n = 204) and LS groups (n = 174), dependent t-test and analysis of covariance (ANCOVA) were used. One-way analysis of variance (ANOVA) and ANCOVA were conducted to compare non-corresponding, one-item corresponding (n = 105) and two-items-or-more corresponding groups (n = 69). Age, body mass index, and dummy variables of medical history of osteoporosis, back pain, hip pain, and knee pain were entered as ANCOVA covariates. The LS group showed significantly poorer physical function according to the dependent t-test, and this poorer function remained significant for the one-leg balance with eyes open performance in both genders and grip strength in men in the ANCOVA (P < 0.05). In the three group comparis