The purpose of this study was to develop and verify the Index of Science Reading Awareness (ISRA) based on a model of an efficient, successful interactive-constructive science reader and three independent metacognitive awareness domains. Several researchers have noted the need for efficient, reliable, and valid measures of metacognition. ISRA data collected on 532 students (Grades 4-8) were analyzed using factor analyses, linear structural modeling, and analyses of variance (ANOVAs) to help verify the model and the test. The factor analyses and linear structural modeling indicated that these data did not support the assumption about the three independent metacognitive awareness domains, but suggested that the model and the test were structured around the design features of science reading, science text, and science reading strategies. One-way ANOVAs indicated significant, predicted reading ability and gender differences but unexpected grade-level differences. The composite metacognitive awareness data indicated that most Grade 4-8 students have surface knowledge about science reading, science text, and science reading strategies, and indicated specific targets for explicit science reading instruction. © 1998 John Wiley & Sons, Inc. J Res Sci Teach 35: 27-51, 1998. Learning from traditional printed and electronic scientific texts has been and will continue to be an important method of achieving and maintaining science literacy. Unfortunately, the processes involved in learning from scientific texts are poorly understood. "The scientist sits down with pencil and paper and slowly works through the article, making notes along the way. Unclear points are pondered over, references are looked up, numerical calculations are checked" (Mallow, 1991, pp. 321-331). The interactive and constructive natures of science reading are apparent as an expert accesses internal and external information sources and actively makes meaning of text. A novice faces a much more demanding task, but the student also accesses prior topical, domain, and task knowledge and works interactively in a sociocultural context to construct understanding from a science textbook (Craig & Yore, 1995, 1996 The interactive-constructive model involving reader, task, and context provides a useful foundation for exploring science reading (Glynn & Muth, 1994;. This model makes central the readers' purpose, prior topic and domain knowledge, beliefs, strategies, language, concurrent experiences, motivation, sociocultural context, and metacognition (Alexander & Kulikowich, 1994;Norris & Phillips, 1994;Ruddell & Unrau, 1994;Holliday, Yore, & Alvermann, 1994). Although the interactive-constructive model acknowledges the importance of social nature of science reading and the need for social interactions, this model assumes that actual learning is a private action not totally explained by the community interactions and grouplevel consensus (McCarthey & Raphael, 1992).The purposes of this study were to clarify a strategic, metacognitive model of an efficient...