Introduction: This study examined the concurrent validity and inter-pair reliability of the proximity detection function on Bluetooth-enabled accelerometers across manufacturer-specified ranges. If valid, this method could be used for objectively measuring shared physical activity participation. Method: Thirty-six dyads aged 21.6 (2.1) years wore two sets of Bluetooth-enabled accelerometers initialized as beacons and receivers to compare accelerometer-measured proximity detection to direct observation under 11 different controlled and free-living conditions. Results: The proportion of minutes in which proximity was detected differed across conditions (p < 0.001), with the highest proportions in the controlled conditions (0.77-1.0). Among the free-living conditions, the different room (0.73 [plus or minus] 0.20) condition had the highest proximity detection proportions. There were no differences in the proximity detection across accelerometer pairs (p = 0.265). Conclusion: Proximity detection with Bluetooth-enabled accelerometers was reliable and had the highest detection rates under the controlled treadmill walking conditions. Despite limitations, using Bluetooth-enabled accelerometers is a promising approach for measuring shared physical activity.