Introduction Electroconvulsive therapy (ECT) is a treatment alternative in bipolar disorder (BD) depression. Cognitive side effects are the major concern limiting its use. Objectives We present data from the Norwegian randomized controlled trial of ECT in treatment resistant depression in bipolar disorder. Aims To compare effects on cognitive function of ECT or algorithm based pharmacological treatment at the end of a six-week acute, BD depression treatment trial. Methods Prospective, randomised controlled multi-centre, six-week acute treatment trial. Pre- and post-treatment assessments with the MATRICS Consensus Cognitive Battery (MCCB); a neuropsychological test battery designed to be sensitive to changes in cognitive function. Sample N = 51 patients ≥ 18 years fulfilling criteria for treatment resistant BD depression (MADRS score ≥ 25). Intervention ECT group: Three sessions per week for up to six weeks, total up to 18 sessions, and right unilateral electrode placement. Algorithm-based pharmacological treatment group: Based on Goodwin & Jamison, 2007. Results Both groups showed a net gain on MCCB scores without significant differences between the study groups. Mean change in MCCB composite T-score was 4.0 (5.7) in the ECT group and 2.7 (3.6) in the pharmacological group (F = 0.78, eta 2 = 0.021, p = 0.383). Conclusion In treatment resistant BD depression ECT and algorithm-based pharmacological treatment have comparable effects on cognitive function assessed with the MATRICS.