Independent ambulation requires adaptability. Self-selected and maximum walking speeds are often both assessed to demonstrate the ability to adapt speed to different tasks and environments. However, purposefully walking at a slow speed (slowWS) could also be an appropriate adaptation in certain situations but has rarely been investigated. The purpose of this study was to assess the reliability, responsiveness, and concurrent validity of slowWS in community-dwelling older adults. This was an observational, cross-sectional study of 110 community-dwelling older adults. Test-retest and inter-rater reliabilities of slowWS were assessed with intra-class correlation coefficients. Standard error of measurement (SEM) and minimal detectable change (MDC95) were calculated to determine responsiveness. Concurrent validity was assessed with Spearman rank-order correlations between slowWS and a battery of tests previously shown to be related to walking speed. Walking speed measurement for slowWS was shown to have excellent test-retest and interrater reliability (ICCs values of 0.971–0.997). Standard error of measurement value was small (0.015 m/sec) and MDC95 was 0.04 m/sec. SlowWS was not found to significantly correlate to any other study variable. Walking speed, whether self-selected, maximum, or slow, can be measured reliably with a stopwatch and specific verbal commands. While slowWS could be beneficial for certain tasks or environments, walking slowly was not associated with age, sex, comorbidity, or measures of cognition, depression, strength, balance, disability, or life-space in this sample. •Slow walking speed measurement has excellent test-retest and interrater reliability.•Slow walking speed SEM and MDC95 values are similar to those for other walking speeds.•Walking at multiple speeds may allow assessment of challenging walking conditions.