The aim of this theoretical study was to evaluate the conceptual model, burden, interpretability, floor or ceiling effects, reliability, validity, and responsiveness of walking speed tests as a measurement of functional limitations in elderly people. A systematic search was conducted in Medline, AgeLine, Embase, CINAHL, and IME, manual searches and references searches. Standardised criteria were applied to assess the quality of the measurement properties. 102 studies were analysed, identifying 18 tests. The most used tests were: 2.44 meters, 4 meters, and 6 meters, carried out at usual gait speed. Most of the findings focused on predictive validity and test-retest reliability; in the latter case, the coefficient values were higher than the quality standards recommended. Scant metric evidence was provided for the attributes burden, interpretability, floor or ceiling effects, and responsiveness. In epidemiological studies, the evidence available supports the use of walking speed tests as predictors of adverse results related with health in elderly people. However, further studies are required to support their viability and applicability in clinical practice, for both screening purposes and to monitor, and evaluate change.