ALMA is now nearing the end of its third cycle of operations, and is transitioning from ‘early science’ to regular PI-driven observing. The array has been operated over the complete range of available baseline lengths, from <10m with the ACA out to the maximum of 16km in the long-baseline configuration. Typically 40 12m-diameter antennas are now used at any one time. In this paper, we summarise the advertised capabilities and how they have evolved in the first 5 years, the proposal pressure and ‘hot spots’, and describe some of the issues with the real measured system performance. We also outline the observing statistics, project completion rates, and papers from ALMA. Finally we highlight some of the new transformational science coming from this facility.