The demonstration benefit from fuel-borne fuel-economy additives to a precision of 1%, or better, traditionally requires very careful experimental design and considerable resource intensity. In practice, the process usually requires the use of well-defined drive cycles (e.g. emission certification cycles HFET, NEDC) in conjunction with environmentally-controlled chassis dynamometer facilities. Against this background, a method has been developed to achieve high-precision fuel economy comparison of gasoline fuels with reduced resource intensity and under arbitrary real-world driving conditions. The method relies upon the inference of instantaneous fuel consumption via the collection of OBD data and the simultaneous estimation of instantaneous engine output from vehicle dynamical behaviour. When applied to a fleet trial employing matched test and control vehicles, relative vehicle fuel economy for multiple fuels may be estimated largely independently of driver, vehicle and environmental factors. The on-road precision achieved by the telematics methodology is shown to approach that obtained in similar fleet trials using conventional chassis dynamometer approaches.