Previous research has led to the development of short batteries of tests that efficiently measure the cognitive functioning of individuals with schizophrenia. To attempt to replicate previous findings, we applied an empirical test selection strategy to archival cognitive test data of two non-overlapping inpatient samples of individuals with schizophrenia or schizoaffective disorder (total N=110). We then extended previous findings by applying the empirical test selection strategy to the archival data of two non-overlapping inpatient samples of individuals with various psychiatric disorders (total N=149). For each sample, tests were selected by examining the relationships between individual test scores and averaged test scores representing global cognitive functioning while taking into account test administration times. Across patient samples, digit symbol coding tasks, verbal fluency tasks, and tests with a processing speed component (Trail Making Test Part A and Stroop) emerged as efficient and effective indicators of overall cognitive functioning. A brief cognitive assessment tool incorporating coding, fluency, and processing speed tasks would provide a valid and clinically useful snapshot of a patient's level of cognitive functioning if more comprehensive testing cannot be completed.