Numerous philosophers accept the differentiation condition, according to which one does not see an object unless one visually differentiates it from its immediate surroundings. This paper, however, sounds a sceptical note. Based on suggestions by Dretske (2007) and Gibson (2002 [1972]), I articulate two ‘principles of occlusion’ and argue that each principle admits of a reading on which it is both plausible and incompatible with the differentiation condition. To resolve the inconsistency, I suggest we abandon the differentiation condition.