Emotional experience is increasingly being measured using experimental tasks, but the stimuli used are often only proxies for the emotion being studied. Stimuli are intended to evoke a distinct emotional experience, but certain designs fail to adequately control for the actual experience in question. In this methodological paper, we review designs used in clinical psychology aimed at measuring emotion and develop the argument of the 'emotion stimulus critique'. Designs of neuroimaging studies on emotion in this context are given preference. We argue that studies often concentrate on standardization of the stimulus material (i.e. words, images, and movies) for eliciting an emotional experience, whereas standardization of the actual participant's experience is seldom performed. Our proposal discusses the use of standardized stimuli in experimental designs and contrasts this with the necessity of controlling for a participant's unique emotional response. We highlight the importance of each participant's 'inner metric', i.e. the individual's experiential anchor, which needs to be taken into account when examining the emotional correlates of psychiatric disorders or psychotherapeutic change. Implications of the emotion stimulus critique for research are discussed within the context of psychology, particularly clinical psychology and psychotherapy.