Fuel use and resupply provisions are vital to all military combat operations, and any significant reduction in the amount of fuel required to sustain the force becomes a large tactical advantage for commanders Developers have been seeking methods that offer even small gains in fuel economy. Small gains for a fleet of thousands of vehicles translate into fewer fuel convoys to theater and large costs savings over time. The challenge to the tester and evaluator is to determine if these small advances are relevant or merely normally expected test variation in the acquired fuel consumption parameters. All too often only the mean fuel economy parameters are compared with and without the new equipment or process without considering test variances inherent in collecting the parametric data. The resulting analysis may then be seriously flawed. Hypothesis testing is a useful statistical method for comparing two sets of test data (sample means and standard deviations) to determine if there is a statistically significant difference between the two sets. Often the two sets of data are made up of small sample sizes (5 test trials are very typical for sets of fuel consumption data). Therefore, for purposes of this discussion, we will consider only hypothesis tests for the differences between two sample means for small (the number of samples is less than 30) sample sizes. Several examples of fictional test data will be subjected to hypothesis testing to show the value of such an approach.