Both in science and in practical affairs we reason by combining facts only inconclusively supported by evidence. Building on an abstract understanding of this process of combination, this book constructs a theory of epistemic probability. It opens with a critique of the well-known Bayesian theory of epistemic probability.