Abstract: The value of evaluations to the public interest depends on their design and utilization. Evaluators from diverse backgrounds present their views on how to improve evaluation research. Case studies and theoretical arguments address 4 areas of concern: obstacles to evaluation; improving evaluation designs; improving measurement; and the role of evaluation. Political forces are seen as major factors limiting the effectiveness of evaluations of federal and state policies and programs. The questions of causal inference and internal validity in evaluation design are examined. Methods for improving measurement are illustrated by studies of mental health policies, social and community programs, job performance, and health technology. The role of evaluation is debated from a methodological and a theoretical perspective. Evaluation's impact on change, and emerging issues for evaluators and evaluation users are discussed. (nm).