© 2015, Royal College of Psychiatrists. All rights reserved. Background: Depression can adversely affect employment status. Aims: To examine whether there is a relative advantage of cognitive therapy or antidepressant medication in improving employment status following treatment, using data from a previously reported trial. Method: Random assignment to cognitive therapy (n = 48) or the selective serotonin reuptake inhibitor paroxetine (n = 93) for 4 months; treatment responders were followed for up to 24 months. Differential effects of treatment on employment status were examined. Results: At the end of 28 months, cognitive therapy led to higher rates of full-time employment (88.9%) than did antidepressant medication among treatment responders (70.8%), χ21 = 5.78, P = 0.02, odds ratio (OR) = 5.66, 95% CI 1.16-27.69. In the shorter-term, the main effect of treatment on employment status was not significant following acute treatment (χ21 = 1.74, P = 0.19, OR = 1.77, 95% CI 0.75-4.17); however, we observed a site x treatment interaction (χ21 = 6.87, P = 0.009) whereby cognitive therapy led to a higher rate of full-time employment at one site but not at the other. Conclusions: Cognitive therapy may produce greater improvements in employment v. medication, particularly over the longer term. Declaration of interest: R.C.S. has served as a consultant for Bristol-Myers Squibb, Cerecor, Cyberonics, Eli Lilly, Forest Pharmaceuticals, Janssen Pharmaceutica, Medtronic, Naurex, Pamlab, Pfizer, Ridge Diagnostics, Shire and Takeda Pharmaceuticals. He has received research and/or grant support from Assurex, Cerecor, Elan Corp, Euthymics Bioscience, Forest Pharmaceuticals, Janssen Pharmaceutica, Jazz Pharmaceuticals, Naurex, Novartis Pharmaceuticals, Otsuka America, Pamlab, Pfizer, Repligen Corp, Ridge Diagnostics, St Jude Medical, and Takeda Pharmaceuticals.