Results of new research into the anomalous results of the 1986 reading portion of the National Assessment of Educational Progress (NAEP) are reported. The original analysis of the 1986 data indicated that the estimated performance level of 9- and 17-year-old students had dropped dramatically since 1984, whereas the performance of 13-year-olds had increased very slightly. Since it was unlikely that such large changes could have taken place in such a short time, the results were not presented to the general public and publication of the results was suspended until further research into their accuracy could be completed or until other corroborating evidence for the declines could be found. To this end, the design of the 1988 NAEP was modified, and new data were collected and analyzed. The main finding from the 1988 data analyses is that the changes in assessment booklets and procedures that were introduced in 1986 had a substantial and unpredictable effect on the estimates of performance. Although many of the same items were used in both the 1984 and 1986 assessments, student performance on these items differed substantially when the items were administered in different contexts. Discovered differences between the 1986 and 1988 assessments were used in a design to estimate the reading performance of students in the 1986 sample. The new estimates show that reading performance declined slightly in 1986 at all age levels as compared to 1984. The declines, however, are not statistically significant. The new data also indicate a rebound in 1988 from the 1986 levels to about the same level of performance exhibited in 1984. The report contains 8 chapters by Beaton, Zwick, Yamamoto and Mislevy detailing the research and analysing the data. Appendices include various types of statistical and methodological information. Twenty-five graphs and 33 data tables are included. (TJH)