To determine the reliability of a standardized protocol by using a hand-held dynamometer (HHD) to measure lower-extremity strength in community-dwelling elderly fallers. Within-session test-retest reliability of the HHD. Balance laboratory of a university. A convenience sample of 41 community-dwelling elders (61-90y) who fell at least once in the previous year. The strengths of 8 lower-extremity muscle groups bilaterally were tested twice, with an intervening rest period of 15 seconds. Maximal isometric force. Test-retest intraclass correlation coefficients (ICCs) were high, generally ranging from.95 to.99 for 1 trial (ICC(2,1)) and from.97 to 1.00 for the mean of 2 trials (ICC(2,2)). There were no significant differences in strength values across trials (P>.05). Men had significantly greater strength than women across all muscle groups (P>.05). Right knee extensor strength demonstrated the largest trial-to-trial difference,.54 kg using a single measurement and.39 kg using the mean of both measurements. Among the 3 lower-extremity muscle groups, the ankle showed higher a coefficient of variation (CV=5.1%-7.4%) than the knee (CV=4.6%-5.1%) or the hip (CV=4.2%-6.3%) when using 1 measure. By using an HHD and a standardized measurement protocol, a novice tester can obtain reliable lower-extremity strength values in community-dwelling elderly fallers. Copyright 2002 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation