Despite their significant contributions to research on self-regulated learning, those favoring online and trace approaches have questioned the use of self-report to assess learners' use of learning strategies. An important rejoinder to such criticisms consists of examining the validity of self-report items. The present study was designed to assess the validity of items to assess 9th-grade students' use of planning, monitoring, and regulation when studying math. To establish response process evidence of construct validity, cognitive interviews were coded to determine whether students' interpretations of the items were consistent with their intended meaning and whether their response choices were congruent with those interpretations. Evidence supported the construct validity of monitoring and regulation items, but to a lesser degree those designed to assess planning. We discuss implications of the evidence for the self-report assessment of learners' use of metacognitive strategies.