OBJECTIVE: To evaluate the effectiveness and efficiency of a tailored multifaceted strategy, delivered by a national clinical effectiveness programme, to implement a guideline on induced abortion. DESIGN: Cluster randomised controlled trial. SETTING AND PARTICIPANTS: All 26 hospital gynaecology units in Scotland providing induced abortion care. INTERVENTION: Following the identification of barriers to guideline implementation, intervention units received a package comprising audit and feedback, unit educational meetings, dissemination of structured case records and promotion of a patient information booklet. Control units received printed guideline summaries alone. MAIN OUTCOME MEASURES: Compliance with five key guideline recommendations (primary outcomes) and compliance with other recommendations, patient satisfaction and costs of the implementation strategy (secondary outcomes). RESULTS: No effect was observed for any key recommendation: appointment with a gynaecologist within five days of referral (odds ratio 0.89; 95% confidence interval 0.50 to 1.58); ascertainment of cervical cytology history (0.93; 0.36 to 2.40); antibiotic prophylaxis or screening for lower genital tract infection (1.70; 0.71 to 5.99); use of misoprostol as an alternative to gemeprost (1.00; 0.27 to 1.77); and offer of contraceptive supplies at discharge (1.11; 0.48 to 2.53). Median pre-intervention compliance was near optimal for antibiotic prophylaxis and misoprostol use. No intervention benefit was observed for any secondary outcome. The intervention costs an average of pound 2607 per gynaecology unit. CONCLUSIONS: The tailored multifaceted strategy was ineffective. This was possibly attributable to high pre-intervention compliance and the limited impact of the strategy on factors outside the perceived control of clinical staff.