Integrating worked examples with problem solving yields more effective and efficient learning, as does intelligent tutoring support for problem solving. This study examines the impact of integrating worked examples and intelligent tutor support for algebra modeling problems. Students in three conditions alternately studied worked examples (either static graphics, interactive graphics or static tables) and solved Algebra Cognitive Tutor problems. A control group solved all the problems with the Cognitive Tutor. Students in the four groups developed equivalent problem-solving skills, but students learned more efficiently in the interleaved worked example conditions, requiring 26% less time to complete the problem set. There were no differences among the four groups in two measures of robust learning a retention test and a transfer test. But students in the static table condition could more accurately describe what algebraic model components represent in problem situations than could students in the other three conditions.