Two experiments are reported which applied signal detection measures to recall and compared them with similar measures of recognition. In both experiments there were more correct responses in recognition but also more false ones. As a result, the signal detection measure of discrimination d' was closely similar for both recognition and recall. The difference between recognition and recall could be accounted for by a more stringent criterion for response in the latter. The second experiment showed that, as in previous experiments on psychophysical discrimination, d' rose approximately linearly with the square root of the number of times the stimuli were repeated. The rate of increase was closely similar for recall and recognition. Possible reasons for the differences of criterion between recall and recognition are briefly discussed. The results emphasize the need to take false as well as correct responses into account when studying either recognition or recall.