Higgins (1987) argued that perceived discrepancies between actual-self and ideal-self increase vulnerability to depression. Ogilvie (1987) argued that discrepancies between actual-self and undesired-self are more powerful determinants of depression. To test these claims, and to assess the effects of mood on reported discrepancies, 190 psychology undergraduates completed measures of their actual, ideal, and undesired selves, along with measures of depression (Depression Adjective Checklist and Beck Depression Inventory) and depression proneness (Depression Proneness Rating Scale). Approximately 3 weeks later, 147 of these participants were assigned to either a positive, negative, or no mood induction procedure, and again completed measures of actual, ideal, and undesired selves. Actual-ideal discrepancies were associated with current depression (' r' =.19, 'p' =.008) and with depression proneness ('r' =.28, 'p' =.001). More importantly, the relationship between actual-ideal discrepancies and depression proneness was, to some extent, independent of current mood, as measured by the DACL (partial ' r' =.19, 'p' =.008). <p>Nevertheless, induced moods also affected the reporting of actual-ideal discrepancies, with negative mood induction increasing discrepancy scores and positive mood induction decreasing discrepancy scores (<math> <f> <g>b</g></f> </math> =.49, 'p' <.0005). The above results are qualified by finding that when induced into a negative mood, high depression-prone participants (those with a history of depression) showed greater increases in actual-ideal discrepancies than low depression-prone participants, suggesting that self-discrepancies are cognitive structures that can be made accessible (primed) by mood. Lastly, no support was found for Ogilvie's claim that actual-undesired self discrepancies can be more powerful determinants of depressive vulnerability than actual-ideal self-discrepancies.