To assess the impact of duloxetine dose escalation on tolerability and efficacy, 516 women with stress urinary incontinence were randomized to receive placebo or duloxetine in one of three regimens: 40 mg BID for 8 weeks, 40 mg QD for 2 weeks escalating to 40 mg BID for 6 weeks or 20 mg BID for 2 weeks escalating to 40 mg BID for 6 weeks. A non-inferiority analysis confirmed that the 20 mg BID starting dose was significantly better than the other two duloxetine regimens for nausea reduction (16.5% vs 25.2% and 29.4%). There were also significant differences in the discontinuation rates (7.5% vs 11.8% and 16.2%). The efficacy after 4 weeks was significantly better with duloxetine than with placebo. Starting duloxetine at 20 mg BID for 2 weeks before increasing to 40 mg BID significantly improved tolerability but did not impact duloxetine efficacy after all the subjects had been on 40 mg BID for at least 2 weeks.