The paper presents a methodology for the Condition Monitoring (CM) of tidal stream turbines. The process is based on the use of, so-called, Transient Monitoring Surfaces (TMS)developed by the authors. In this paper the TMSs have been used to detect rotor imbalance faults. To test the use of TMS for CM drive train simulations were undertaken. The simulations undertaken relate to a lab-scale turbine subjected to turbulent flows and an optimal � control scheme based on Vector Oriented Control (VOC). The simulations are parametrised based on experimental data relating to testing undertaken with varying degrees of rotor imbalance. Use of the TMS gave promising results for the detection of various rotor imbalance conditions. Differing levels of discrepancies between the ’normal operating’or ’baseline’ surface were found for differing fault severities. It was also found that a minimum amount of data is required to gain convergence in the surface structure - in this case data sets relating to 5 rotations of the turbine were required to make a suitable fault detection.