The　procedure　for　self-regulated　learning　ability　evaluation　inevitably　involves　human　being＇s　subjective　judgments　and　uncertainties.　In　order　to　model　these　uncertainties,　this　paper　presents　an　interval　evidential　reasoning　(IER)　approach　for　autonomous　learning　ability　assessment,　which　uses　the　analytical　evidential　reasoning　(ER)　algorithm　to　aggregate　all　evidence　simultaneously.　Two　pairs　of　nonlinear　optimization　models　are　employed　to　estimate　the　upper　and　lower　bounds　of　the　combined　belief　degrees　and　to　compute　the　maximum　and　the　minimum　expected　utilities.　The　numerical　example　gives　an　application　of　IER,　which　demonstrates　its　validity　for　college　students＇　autonomous　learning　ability　evaluation.