Objective Meta‐analyses of antidepressant drug treatment trials have found that increasing age is associated with a less favourable outcome. Because the prevalence of geriatric characteristics, like disability, medical co‐morbidity, malnutrition, cognitive (dys)function and frailty increase with age and are associated with depression, these characteristics are likely to modify the treatment outcome of antidepressant drugs in late‐life depression. This review examines how these five characteristics are taken into account in randomised controlled trials (RCTs) with antidepressant drugs for major depressive disorder in patients aged 60 years or above. Design A systematic search in PubMED, PsychInfo and EMBASE, from the year 2000 onwards, yielded 27 RCTs, with a total of 6356 subjects with a median age of 71 years. Two reviewers independently assessed whether each characteristic was considered as inclusion or exclusion criterion, descriptive variable, stratification variable, co‐variable, outcome measure, or in adverse effect monitoring. Results Malnutrition and frailty were not taken into account in any study. Disability was used as an outcome measure in five studies. Two studies explicitly included a population with possibly serious medical co‐morbidity. Cognitive status was the only condition taken into account as co‐variable (n = 3) or stratifying variable (n = 1) and was used as outcome measure in seven studies. Conclusions We conclude that geriatric characteristics are rarely taken into account in RCTs on antidepressant drugs in late‐life depression, and studies including the oldest adults are underrepresented. This warrants recruitment of the oldest adults and adjustment of treatment strategies in future studies. Copyright © 2016 John Wiley & Sons, Ltd.