AbstractA 6-week experimental program was written to teach six basic spelling principles to sixteen sixth-grade classes. The classes were randomly assigned to four experimental treatment groups-each of which used a different combination of method and procedure. One immediate and one delayed posttest were given to determine spelling achievement in terms of retention, transfer, and problem-solving abilities. Analysis of variance procedures used to analyze the results showed no significant differences for method, procedure, or interaction. Several of the results approached significance, however, as the discover)’ method and the teacher-guided group procedure showed some advantage on the delayed transfer test. Implications for further research are suggested.