Previous studies have shown that a change in an existing object is not as effective in capturing attention as the appearance of a new object. This view was recently challenged by Lu and Zhou (Psychonomic Bulletin and Review 12:567–572, 2005), who found strong capture effects for an object changing its color. We suspected that this finding critically depends on a procedural particularity in Lu and Zhou’s study, namely that the color of the unique item and the color of the no-unique items randomly switched between trials. In the current study we replicate Lu and Zhou’s capture effect (Experiment 1) and show that no capture occurs when the color-to-stimuli assignment is fixed (Experiment 2). Two further experiments suggest that the capture effect in Experiment 1 is not because the unique item switched color (Experiment 3), but because all the no-unique items switched color (Experiment 4). The results are discussed considering top-down modulation and inter-trial priming effects.