Purpose:To describe differences in test-retest reliability of balance measures produced by the Biodex Balance System.Methods:Subjects performed balance tests in 4 studies, each involving two 30-second tests under varying conditions. Test-retest reliability coefficients were calculated for the overall stability index (SI) under each condition. Test-retest reliability was computed for subjects' limits of stability (LOS).Results:Intraclass correlation coefficients (ICC3,1) ranged from .44 to .89 for static balance tests; variability between the 2 tests ranged from 21% to 37%. Tests at lower stability levels or without visual input produced the highest ICCs. ICCs for the LOS tests ranged from .64 to .89. These measures were less variable (6% to 8%) than the static measures of balance.Conclusions:Test-retest reliability of the SI produced by the Biodex Balance System is acceptable for clinical testing and is comparable to other balance measures currently in use.