The link between instruction in middle school science and assessment in ninth grade science in a small rural midwestern school district was investigated. Science students, in middle school grades 6-8 (917) and their 13 science teachers responded to surveys used to characterize science learning opportunities in the middle school classes. Ninth grade students (309) responded to a set of open-ended science questions that were developed from a standardized science test. The survey data provided a contextual framework for interpretation of ninth grade student performance on the open-ended science questions. The open-ended questions included questions for which students were required to graph and interpret data, write conclusions, identify control variables, and judge the validity of information. These open-ended questions were scored by three raters according to a scoring rubric. Middle school students did report having science learning opportunities to practice science as inquiry, and their teachers reported that much of the science inquiry practice was contextual. At least 90% of the ninth graders were from the same school district, but when practice in controlling data and making predictions was reported, student data from the open-ended questions suggested minimal transfer of these skills. The interpretation and use of graphic information and the actual graphing of data were areas that were problematic for ninth graders, as was the ability to provide reasons in support of answers. Recommendations based on this study include using a variety of contextual settings for framing science inquiry practice and more attention on discussing, reading, and writing in the content area. (Contains 8 tables and 12 references.) (Author/SLD)