In the present paper and the discussion that follows, the author presents aspects of test construction and a careful description of instructional objectives. Constructing tests involves several stages such as describing language objectives, selecting appropriate test task, devising and assembling test tasks, and devising a scoring system for evaluation purposes. Tests can be classified according to their reference, that is, something external that serves to give test scores meaning. Thus, the author distinguishes between norm referenced and domain referenced testing. The former are in reference to performance and the latter to a general domain of skills and knowledge. In other instances, tests can derive significance through reference to more restricted domains such as those delineated in instructional plans and practices. These are called objective referenced tests. The focus here is on objective referenced tests because of their particular relevance to evaluation in second language classroom. The author also presents guidelines for devising closed-ended and open-ended test tasks. These guidelines are part of a larger process of devising valid tests that are compatible with the focus, range, and standards of performance specified or included in instructional objectives. [For complete volume, see ED567118.]