This recommended practice provides a standardized test procedure for comparing the in-service fuel consumption of two conditions of a test vehicle or of one test vehicle to another when it is not possible to run the two or more test vehicles simultaneously. An unchanging control vehicle is run in tandem with the test vehicle(s) to provide reference fuel consumption data. This procedure is especially suitable for testing components which require substantial time for removal and replacement or modification, such as engines, transmissions, tag-axles, and cab sheet metal. This procedure may also be used for comparison of entire vehicles and for easy-to-change components (those referenced in the Type I test described in SAE Recommended Practice, SAE J1264). The test may utilize fleet vehicles operating over representative routes. The result of a test using this procedure is the percent difference in fuel consumption between the two test vehicles or the difference in fuel consumption of one vehicle in two different test conditions. The fuel measurement method is a key factor in determining the overall accuracy achievable with this procedure. If the weighing method is used, overall test accuracy is best and, based on test experience, will be within +1% (for example, 6% measurement improvement can be from 5–7% actual improvement). (See Section 7, Test Accuracy.) The following four basic rules must be applied to this procedure to insure test result validity:

1. a The test routes and cargo weight should be representative of actual operation.
2. b A single test is inconclusive regardless of the results. A single test should be taken as an indicator. Test results must be repeatable to have validity.
3. c The more variables controlled, the more conclusive the results.
4. d All test procedures or methods are accurate within prescribed limits.

If the component or system being tested by a given procedure shows a degree of improvement which is equal to or less than the accuracy limit of the procedure, an additional number of tests should be conducted to determine its true value. If a number of such tests do not show consistent results, then one must conclude that the changes caused by the component or vehicle system are less than can be measured by the test procedure.