This review analyzes traditional standardized assessment practices to determine if they recognize cultural, linguistic, socioeconomic, and disability differences. It then examines the extent to which these assessment practices are being used in transition for decision-making and predicting post-school outcomes. The review begins with a historical overview of testing. It discusses three types of standardized assessments: norm-referenced, minimum-competency, and criterion-referenced tests. Alternatives to traditional standardized assessments are explored. The review concludes that the failure to consider sociocultural milieu, socioeconomic status, and linguistic and disability differences in test construction leads to false information about the status of learning and compounds bias in testing. Thus, minorities and students with disabilities, in particular, suffer as a result of traditional assessment practices, which have proven to be inaccurate and inconsistent, yet continue to be used in prediction, decision-making, and inferences about student performance and lifelong success. Current assessment practices stifle the post-school success of minorities and students with disabilities due to their inappropriate norming standards, discrimination, and exclusion. By examining individual student skills on specific tasks, alternative assessments appear to be a solution to eradicating biases. Reasonable and limited use of traditional standardized assessments for the purpose of decision-making, prediction, and inferences about students' post-school activities is recommended. (Contains 42 references.) (JDD)