Driver behaviour can strongly affect fuel consumption, and driver training in eco-driving techniques has been shown to reduce fuel consumption by 10% on average. However the effects of this training can be short-lived, so there is an apparent need for continuous monitoring of driver behaviour. This study presents a driver advisory tool which encourages eco-driving, and its evaluation in the field. The system, developed by Ashwoods Automotive Ltd (UK) and the University of Bath (UK), is aimed at fleet operators of light commercial vehicles, where the driver is typically a company employee. A significant strength of the system is that it has been designed for easy integration with the vehicle CAN-bus, reducing complexity and cost. By considering the Inertial Power Surrogate (speed times acceleration) the core algorithm is able to identify behaviour which is likely to increase fuel consumption. The algorithm also enforces the advice of the Gear Shift Indicator to encourage earlier upshifting. Fuel saving is therefore encouraged both by reducing the power demand of the drive cycle and by operating the engine in a more efficient region. Instantaneous visual feedback to the driver is augmented by audible warnings, which avoids the need for the driver to concentrate on the display, reducing cognitive loading. Reports on driver performance are made available to the fleet manager so that fuel efficient and safe driving can be encouraged.Trials of the system ran for four weeks and included 15 vehicles from 7 companies; a total of 39 300 km of trip data were collected during more than 1 100 hours of real world driving. Results show an average fleet reduction in fuel consumption of 7.6%, with some vans showing savings up to 12%. Furthermore, reductions of over 10% in both average engine speed and throttle activation were observed. It was also noted that the system shows strong potential for synergetic use with a hybrid electric powertrain, where modifying driver behaviour may facilitate energy harvesting from regenerative braking.