Designing a test using three-parameter item response theory (IRT) is discussed. A brief review of IRT is followed by a discussion of two types of test design: (1) selecting items using confidence envelopes (confidence envelope method); and (2) using item characteristic curves and their confidence intervals (test envelope method). The confidence envelope method and the test envelope method are evaluated based on their reliability coefficients, using a set of seven items. Results illustrate that the test envelope method, in which optimum ability levels are matched from an item bank, should result in a more reliable test. Six tables and three graphs illustrate the analysis. (SLD)