The problem of determining optimal test lengths with fixed total testing time has proved to be a difficult one for criterion-referenced test developers. An algorithm is needed which can be used by test developers to allocate available testing time to maximize the validity of their total criterion-referenced tests or testing programs. To be maximally useful, the algorithm should allow test developers to specify: (1) the various informational needs they have in relation to the objectives measured by a test or the uses they have for several tests in their testing program; and (2) the relative importance that they attach to their informational needs. This paper describes such an algorithm for determining the number of items to measure each objective in a criterion-referenced test when testing time is fixed and when the objectives, in general, vary in their levels of importance, reliability, and validity. Four examples in the paper highlight possible applications of the procedure. While the offered solution is simple to apply, it is only applicable when basic psychometric data are available on the scores at the objective level and when the objective scores are being used to make descriptive statements about examinee performance. (PN)