This thesis investigated the role of valence in memory updating in depressive rumination. In two paradigms—a novel working-memory updating task and a continued-influence paradigm—depressive ruminators demonstrated no deficit in updating negative information. Findings are inconsistent with the literature, which has relied purely on arguably inadequate response-time and list-recall measures. However, the observed pattern of behaviours was consistent with negative attentional bias—a well-established phenomenon in this population. Results from the continued-influence paradigm in particular suggest implications for clinical practice, as this paradigm has strong ecological validity and thus better reflects real-world situations involving conceptual updating.