The United States Court Interpreters Act (US Congress, 1978) requires that interpreters in US federal courts be certified through a criterion-referenced performance test. The Federal Court Interpreter Certification Examination (FCICE) is a two-phase certification battery for federal court interpreters. Phase I is a multiple-choice Written Examination (WE) used to screen candidates for eligibility to take the Phase II criterion-referenced Oral Examination (OE). This study evaluates the predictive validity of the WE in relationship to examinee performance on the OE, using an experimental sample of examinees who took the tests concurrently. Results indicate that the WE correlates appropriately with performance on the OE (0.648 and 0.676 for the English and Spanish subtests respectively). Nearly all of those who failed the WE also failed the OE, although several false negatives were identified. However, when the data were reanalysed using a lower cut score for the Phase I WE, nearly all false negatives were eliminated. A follow-up study replicated the findings with a different sample. The authors discuss the advantages and disadvantages of lowering the passing score on the screening test in light of their findings. The study and subsequent discussion provide an example of how an established passing score can be reconsidered and modified.