Cognitive errors (CEs) are evidenced to be related to depressive thinking in major depressive disorder (Beck Et Al., 1979; Dozois & Beck, 2008). Studies using self-report questionnaires demonstrate that CEs are more prevalent in individuals with depression than in non-depressed individuals (Gupta & Kar, 2008) and that CEs are related to depression severity (Miranda & Mennin, 2007). The study discussed in this paper aimed to describe CEs in depressed patients and examined the relationship between CEs and severity of depression. Participants (N = 45) undergoing cognitive therapy were assessed for CEs and for depression at session three using the Cognitive Errors Rating System (CERS; Drapeau et al., 2008) and the Beck Depression Inventory (BDI; Beck et al., 1979). Participants had more negative CEs than positive, and the most prevalent cluster of CEs was selective abstraction. Participants deemed as being "high distorters" on the CERS had significantly more negative CEs, but not positive CEs, than "low distorters" despite not differing on BDI scores. Psychotherapy research and practice implications are discussed.