The measurement of power performance is an important procedure in the design verification and ongoing health monitoring of a tidal turbine. Standardised methods state that the performance should be measured relative to two independently located flow sensors, the arrangement of which is often non-trivial and necessitates additional cost. Recent interest in the usage of flow sensors mounted on the turbine has demonstrated their capabilities in profiling the rotor approach flow, but this instrument configuration is not recognised in the performance assessment standard. This study evaluates the merits of the turbine mounted configuration by measuring the performance of a tidal turbine relative to this reference and to a conventional seabed placed instrument. The turbine mounted sensor is found to provide a better reference of the free-stream conditions, evident from an improved agreement with theoretical predictions of device performance and a reduced amount of variation in the results. This new method could reduce both the costs and uncertainty associated with existing performance assessment best practices.