Educational systems promote student growth in a variety of dimensions. Basic content knowledge can be effectively assessed with multiple-choice and completion tests. However educational reforms have become more concerned with higher-order cognitive dimensions (problem-solving, creativity), social dimensions (communication skills, ability to work in groups) and other dimensions (life-long learning). While they are objective and efficient, traditional assessment measures may not serve these kinds of goals as well as other types of measures. Before we can choose an accurate, efficient method of assessment, we must clearly understand the goals of science instruction. Do these goals encompass only the basic memorization of facts? If so, our traditional methods may be sufficient. If we wish to institute a science program that encourages dimensions that go beyond these basics, we will need to develop a repertoire of additional assessment methods. The organization of this manual is intended to aid teachers in developing expertise in identifying learning goals, choosing assessment methods, and communicating assessment results in such a way that student performance is enhanced. [This report was produced by the SERVE Center at the University of North Carolina at Greensboro. For previous edition, see ED363622.]