Abstract Acoustic Doppler current profilers (ADCPs) are commonly used to assess mean currents and turbulence at energetic sites. Since 2014, five-beam ADCP configurations have become more common, but conventional analysis of turbulence properties is still based on the four-beam Janus configuration. We use measurements from a single site to investigate improved estimates of turbulent kinetic energy (TKE) that are made possible by the addition of a fifth vertical beam. We conclude that four-beam estimates of TKE are suitable in most cases, and exhibit lower variance than five-beam estimates, but are more prone to contamination by wave activity.