This report presents the methodology and findings of a national survey of the performance of approximately 9,000 11-year-old children in science. Performance was assessed in six categories: (1) symbolic representation (reading information from graphs, tables, and charts); (2) use of apparatus and measuring instruments; (3) observation; (4) interpretation and application; (5) design of investigations; and (6) performance of investigations. In addition, findings related to the nature of and provision for science activities in schools are presented. Overall patterns found were: a higher performance with familiar situations for "observation"; good performance for "symbolic representation" but a wide disparity between forms of such representation; generally good performance for "using apparatus and measuring instruments"; generally good use but often accompanied by poor explanations for "interpreting application"; generalized and partial answers, the most satisfactory types received, for "designing investigations"; and a context-dependent pattern for "performance of investigations." Data on sex differences for these and other categories are also presented and discussed. (Included in appendices are major findings from a similar survey conducted in 1980, the questionnaire used to obtain data on school science activities, and other supporting documentation.) (JN)