Abstract: Vessel operating data is collected and stored while training a monitoring system. The training process includes navigating along a first path while running at a constant RPM for each of multiple RPM settings. The step is repeated for an opposite heading. Data from one or more training runs is used to derive models of estimated vessel performance. A calibration process may be performed to determine performance degradation due to marine growth and/or propeller damage. During real time operation, vessel performance is monitored and displayed. An operator may compare actual performance versus estimated performance in generally real time using the training data and real time data. The operator may change course or RPM to improve performance, such as to improve vessel miles per gallon.