The analysis of routines of older adults is a growing field of research, since detecting changes in behaviour patterns can be an indicator of physical or cognitive decline. This requires continuous monitoring of people over long periods. This work introduces a novel dataset, publicly available for the scientific community, which is composed of the BLE RSSI (Received Signal Strength Indicator) measurements recorded by the smartwatches that the elderly volunteers wore, all of them living in a nursing home where a set of BLE beacons have been deployed. The beacons were deployed in different rooms of the building (bedrooms, gym, restaurant, etc.), where the volunteers spend most of the time. This information can be beneficial to obtain, for instance, a symbolic location of the monitored people or to study their routines. The activities that the volunteers perform throughout the day are reported by their caregivers. In addition, the dataset includes ground truth measurement information for the rooms in which the BLE beacons have been installed.