Five experiments demonstrate that when dots appear beside a briefly presented target object, and persist on view longer than the target, the flanked object is perceptually altered by the dots. Three methods are used to explore this "object trimming effect". Experiments 1-3 assess participants' conscious reports of trimmed digits, Experiment 4 uses repetition priming to explore the target representation, and Experiment 5 examines the perception of apparent motion in trimmed targets. Results of all three methods indicate that object trimming is influenced by mechanisms of perceptual grouping that operate on target representations prior to conscious access. Separate contributions from visual crowding and backward masking are also identified. These results imply that common-onset masking does not always result from the target being substituted by the mask, but that target and mask can sometimes maintain separate mental representations. (Contains 4 footnotes, 7 tables, and 10 figures.)