The aim of this thesis is to consider what it means to say that two risks are incommensurable and to explore if incomparability - comparison failure with respct to a common ordinal scale - ever occurs, and how to model it if it does. Essay I is a critical examination of the most prominent argument for incomparability, the so-called small improvement argument (SIA). It is argued that the argument fails because it conflates incomparability and a kind of evaluative indeterminacy. Essay II outlines so-called margin of error principles for comparative value judgements. They are based on the idea that if a proposition concerning the value relation between two value-bearing options is true, but there are sufficiently similar cases in which it is false, it is not available to be known. The usefulness of these principles is demonstrated by utilizing them in an epistemological case against SIA. Essay III presents a novel account of incomplete preference orderings which acknowledges that incomparability can vary in degrees. This is achieved by means of a probabilistic analysis of preferences.