Background Reactive agility tests have become popular in sport for talent identification; however, the ability of these tests to evaluate physical function and falls risk in clinical populations warrants future study. Aim To examine the reliability and construct validity of a novel functional reactive agility test (FRAT) across the lifespan. Methods Forty-three young (24 ± 2 years), 32 middle-aged (50 ± 2 years), and 19 old (66 ± 4 years) men performed a FRAT that included a rapid lateral movement (1.5 m) in response to a randomly delayed visual stimulus. Test–retest reliability and minimum difference (MD) scores were determined from a subset of participants. Results There was no systematic error (P > 0.087) between testing days, and the intraclass correlation coefficients, standard error of measurement (% of the mean), and MD values for decision time, movement time, and total time ranged from 0.876 to 0.949, 4.16–9.24%, and 0.128–0.138 s, respectively. The young men had faster decision times (P = 0.027) when compared to the older men and faster total times when compared to the middle-aged and older men (P < 0.001). Discussion The FRAT demonstrated acceptable reliability and construct validity between different age groups. Conclusion Due to its limited space requirements, the FRAT may serve as a useful tool in future studies examining clinical populations.