An economical method for reliably determining fleet fuel savings that arises out of the application of a vehicle fuel saving measure is presented. The method relies on testing the device on a random sample of vehicles during normal fleet operations. The sample size is determined by variations in fuel economy inherent in the fleet and the minimum increase in fuel economy needed for a desired return on investment in the device. This reduces the probability that the sample is biased. The sample fuel economy is compared to that of an untreated control group of trucks which can be the rest of the fleet or a subset with certain characteristics, like similar idling percentage. The final results provide a range of fuel economy enhancements at a given confidence level. The method has been applied to validate in operation fuel economy increase of a flatbed tractor trailer fleet due to Sinhatech's Deturbulator aero-drag reducing device.