Dose escalation is often used in depressed patients who fail to respond to standard doses of selective serotonin reuptake inhibitors, but clinical efficacy is equivocal. We aimed to reassess the efficacy of paroxetine dose escalation and quantify whether paroxetine dose escalation increases occupancy of the serotonin transporter (SERT) more than placebo dose escalation in a randomized controlled trial. We recruited 107 nonpsychotic, unipolar depressed outpatients (18-70 years; Hamilton Depression Rating Scale (HDRS(17)) >18) from primary care and psychiatric outpatient departments. After 6 weeks, open-label paroxetine 20 mg per day (T0), nonresponding patients (HDRS(17) decrease <50%; n=60) were randomized to double-blind paroxetine (30-50 mg per day as tolerable) or placebo dose escalation (paroxetine 20 mg per day+placebo). Patients were followed until 6 weeks after randomization (T1). Forty-nine patients, drug free at study entry, underwent single-photon emission-computed tomography (SPECT) scanning before treatment and were scanned repeatedly at T0 and T1. Paroxetine serum concentrations and SERT occupancy were determined at T0 and T1 (n=32). We terminated the dose-escalation trial after an interim analysis. Thirty nonresponding patients were randomized to paroxetine (46.7+/-5.5 mg per day), 27 to placebo dose escalation. Response rates were 10/30 (33.3%) and 10/27 (37.0%), respectively. Repeated measurement analyses showed no significant effect for treatment (p=0.88, exceeding a priori stopping rules for futility (p>0.5)). Overall dropout was higher for placebo (26.7%) than paroxetine (3.3%; p=0.03). Paroxetine dose escalation increased paroxetine serum concentrations (p<0.001). SPECT measurements (12 patients randomized to paroxetine (46.9+/-4.8 mg) and 14 to placebo dose escalation) showed no significant increase of midbrain SERT occupancy (2.5+/-26.4%, paroxetine; 3.1+/-25.8% placebo; p=0.687) nor in diencephalon (p=0.529). Paroxetine dose escalation in depressed patients has no clinical benefit over placebo dose escalation. This is explained by the absence of significant increases of SERT occupancy by paroxetine dose escalation, despite increased paroxetine serum concentrations (ISRCTN44111488).