This paper is intended to guide developers of wireless systems who are puzzled by the vast number of radio configuration parameters and options. We provide experimental data comparing power consumption of Bluetooth Low Energy (BLE), ZigBee and ANT protocols for a cyclic sleep scenario, in which a short-range and low-power wireless sensor node periodically sends a data packet to a remote ‘hub’ with intervening sleep intervals. Devices such as wearable health monitors often use this scenario to interface with a mobile phone-based hub. For all measured sleep intervals BLE achieved lower power consumption (10.1 uA, at 120 s interval), compared with ZigBee (15.7 uA), and ANT (28.2 uA). Power consumption differences can be attributed to the time taken for a node to connect to the hub after waking up and the use of sleep between individual RF packets. For the three protocols we determined a sleep interval at which the trade-off between power consumption and data rate is optimized.