Despite evidence of negative psychological sequelae and unmet needs, there are few evaluated interventions for informal caregivers in cancer and palliative care. The aim of this article is to debate the strengths and limitations of randomized controlled trials (RCTs) and other designs that can be used to evaluate the effectiveness of these interventions. Psycho-educational interventions are used as example for this debate article, as a number of studies of various designs evaluating this type of intervention have been published. Systematic searching in Medline and the bibliography of a relevant systematic review identified five RCTs, one pre-test/post-test study with a control group and six one-group pre-test/post-test studies of psycho-educational interventions for caregivers. The methodological strengths and weaknesses were assessed. RCTs are seen as the gold standard, but can have important limitations in the context of carer intervention research, including biased recruitment and low generalizability, problems with blinding and attrition. Pre-test/post-test studies with a control group may be more feasible and more generalizable. Their crucial limitation is selection bias. Before–after studies are compromised by additional specific biases and therefore are the weakest of all discussed designs. After analysing the strengths and weaknesses of the mentioned study designs, this paper presents strategies to address the limitations of RCTs evaluating psycho-educational interventions for carers in cancer or palliative care.