Abstract
To investigate the cognitive functions of underlining, college students studied and reviewed a history chapter under one of three conditions: underlining during study and reviewing the underlined text, underlining when studying but reviewing clean text, and studying and reviewing without underlining. Scores on an exam measuring recognition of facts and inferences indicated that subjects who underlined when studying and then reviewed their underlined chapter scored significantly lower on inferential recall than the other two groups. Results indicated that underlining apparently does not serve an encoding or review function, and may be counterproductive for inferential recall. Results are discussed in terms of students’ self‐reported use of underlining.