<p>In the new mechanical philosophy, a mechanism is said to comprise entities and activities, whose organization gives rise to a phenomenon. What lies outside the boundary of a mechanism is typically treated as irrelevant to the phenomenon and is regarded either as a background condition or as noise. I argue, however, that noise may sometimes contribute to the emergence of a phenomenon. My aim is not to claim that all or most instances of noise invariably contribute to a phenomenon, but rather to explicate that certain relevant forms of noise can play a contributory role. The implication of this study is that a phenomenon may not be determined solely by the respective mechanism alone. Instead, it may be shaped, to some extent, by noise alongside the mechanism’s entities and activities. Without taking noise into account, some phenomena may not be satisfactorily explained. Finally, I argue that noise is <i>prima facie</i> a constituent of certain mechanisms, and that its ontological status warrants further investigation.</p>