In a recent study Diamond, Kirkham and Amso (2002) obtained evidence consistent with the claim that the day-night task requires inhibition because the picture and its corresponding conflicting response are semantically related. In their study children responded more accurately in a dog-pig condition (see /day picture/ say "dog"; see /night picture/ say "pig") than the standard day-night condition (see /day picture/ say "night"; see /night picture/ say "day"). However, there is another effect that may have made the day-night condition harder than the dog-pig condition: the response set effect. In the day-night condition the names of the two stimuli ("day" and "night") and the two corresponding conflicting responses ("night" and "day") are from the same response set: both "day" and "night". In the dog-pig condition the names of the stimuli ("day", "night") and the corresponding responses ("dog", "pig") are from a different response set. In two experiments (Experiment 1 with 4-year-olds (n = 25); Experiment 2 with 3 1/2-, 4-, 5-, 7- and 11-year-olds (n = 81)) children were tested on four experimental conditions that enabled the effects of semantics and response set to be separated. Overall, our data suggest that response set is a major factor in creating the inhibitory demands of the day-night task in children of all ages. Results are discussed in relation to other inhibitory tasks.