In her inceptive study of promising, Astington (1988) found that five-year-olds failed to distinguish promises from predictions when judging story characters' utterances. Instead, their responses were based on whether the promised event had actually occurred. There is reason to believe, however, that the children's poor performance was an artefact of Astington's procedures. We tested this possibility with 32 children (mean age 5;8: range 5;7-6;2) by including several variations of Astington's procedures, but found no effects on the pattern of results. This suggests that Astington's findings are an accurate reflection of the developmental course of knowledge about promising.