There is little evidence regarding which therapist strategies promote cognitive change in cognitive therapy (CT) of depression. Drawing from a sample of CT patients, we selected two consecutive sessions for which patients reported markedly different amounts of cognitive change (CC; i.e., a low and high cognitive change session). We then investigated whether four observer-rated psychotherapy process variables differentiated high and low CC sessions. Our analyses focused on 62 patients with large session-to-session differences in self-reported CC. Results from single predictor models showed the therapeutic alliance and therapists’ use of cognitive methods predicted high versus low CC session type. In a model including multiple predictors, only cognitive methods remained significant. These findings are consistent with the view that cognitive methods promote CC in CT.