This article examines the stability of objective chance. I defend the stable chance thesis (SCT): that in any given possible world, any pair of intrinsic duplicate physical setups with the same chances of being subject to the same external influences must yield the same chances. I argue that SCT compares favourably to rivals in the literature. I then consider a challenge to SCT involving time travel and causal loops. I argue that SCT survives this challenge, but that such cases expose chance as less stable than we might unreflectively have thought. In particular, the chances associated with a physical setup are sensitive to the way in which that setup is embedded in a wider causal structure.