Background: Wrist activity trackers (WATs) are becoming popular and widely used for the monitoring of physical activity. However, the validity of many WATs in measuring steps remains unknown. Objective: To determine the validity of the following WATs: Garmin Vívofit (Vívofit) and Polar Loop (Loop), by comparing them with well-validated devices, Yamax Digiwalker SW-701 pedometer (Yamax) and hip-mounted ActiGraph GT3X+ accelerometer (ActiGraph), in healthy adults. Methods: In free-living conditions, adult volunteers (N = 20) aged 25 to 52 years wore two WATs (Vívofit and Loop) with Yamax and ActiGraph simultaneously over a 7 day period. The validity of Vívofit and Loop was assessed by comparing each device with the Yamax and ActiGraph, using a paired samples t-test, mean absolute percentage errors, intraclass correlation coefficients (ICC) and Bland-Altman plots. Results: The differences between average steps per day were significant for all devices, except the difference between Vívofit and Yamax (p = .06; d = 0.2). Compared with Yamax and ActiGraph, the mean absolute percentage errors of Vívofit were -4.0% and 12.5%, respectively. For Loop the mean absolute percentage error was 8.9% compared with Yamax and 28.0% compared with ActiGraph. Vívofit showed a very strong correlation with both Yamax and ActiGraph (ICC = .89). Loop showed a very strong correlation with Yamax (ICC = .89) and a strong correlation with ActiGraph (ICC = .70). Conclusions: Vívofit showed higher validity than Loop in measuring daily step counts in free-living conditions. Loop appears to overestimate the daily number of steps in individuals who take more steps during a day.