Abstract The functional response of parasitoids is a main component of models that describe host–parasitoid interactions. When the functional response assumed in a model (e.g., a population model) is inaccurate, predictions by the model based on the functional response also become unreliable. This study reports that conventional functional response models may commonly fail to capture the actual functional response of parasitoids and discuss its implications. To assist the discussion, an experiment was conducted to describe interactions between the pupal parasitoid Dirhinus giffardii (Hymenoptera: Chalcididae) and its host Bactrocera dorsalis (Diptera: Tephritidae) in the laboratory. Results showed that the parasitoids typically interacted with all hosts in the environment but parasitised a small fraction of them (e.g., most encountered hosts were rejected), which deviates from the assumption of mechanistic functional response models. An increasing number of published studies on functional response also suggest (although circumstantially) the rejection of unparasitised hosts is common, suggesting that this behaviour is widespread among parasitoid species. By using the experimental results, the study discusses why existing hypotheses are insufficient for the observed behaviour and the implications of this behaviour within the realm of theories surrounding host–parasitoid interactions.