In conditional reasoning tests (CRTs), participants are asked to choose a response that follows logically from premises presented in the item stem. CRTs rest on the rationale that individuals with a high standing on an undesirable trait will draw inferences from the set of premises that provide a justification for their trait standing. In this preregistered study, we presented the CRT for aggression either with or without premises (N = 778). In line with CRT theory, results revealed that CRT-A scores were affected by the omission of premises. On the item level, for six of 22 items, the frequency of choosing illogical responses did not differ when the premises were removed. Correlations with self-reported aggression were similar for both versions of the CRT-A. Thus, findings speak to the theorizing behind CRTs, but also point to possible improvements on the item level.