Metaepistemology is the study of the underlying assumptions of epistemology. As the theory of knowledge, epistemology is concerned with questions about what knowledge is and how much people can know. Metaepistemology, by contrast, investigates what the aims and methods of epistemology should be, whether there are objective facts about what people know, and related issues.

There are differing views in metaepistemology about the nature and methods of epistemology. Epistemology is usually seen as a field that evaluates what the right things to believe are and prescribes how one ought to form beliefs. Traditional characterisations emphasise the use of reflective thought and intuitions rather than empirical evidence. Other views include the idea that epistemology should use methods more similar to the sciences—like experiments—or that it should focus on the practical impact of the concepts it employs. Feminist epistemology has extended alternative views like these to criticise gendered bias in epistemology.

Metaepistemology investigates epistemic facts, like facts about what people know. According to epistemic realists, facts about knowledge are objective and depend on the way the world is rather than subjective opinion. Anti-realists deny that there are such facts, either by denying their existence altogether or by denying that they are objective. For example, expressivists argue that statements about knowledge do not represent facts, but express attitudes like "this belief is good enough". Views that attempt to find a middle-ground between realism and anti-realism include quasi-realism and constitutivism. Metaepistemology also investigates how it is possible to know about epistemic facts, an area called the epistemology of epistemology.

As a discipline, epistemology does not describe what people actually do believe, it shows what people should believe or what they have justification to believe. Some epistemologists, for instance, assert that everyone has an obligation to only hold beliefs based on evidence. This is a feature known as normativity and it leads to a number of questions in metaepistemology. For example, it is disputed whether people can choose what to believe and whether judgements about evidence can motivate people to believe the right things. Other questions include why and how true beliefs are valuable. Since these problems resemble some issues discussed in the field of metaethics, metaepistemologists examine the similarities and differences between the two disciplines.

Metaepistemology is the branch of epistemology focused on the fundamental assumptions of epistemology. As a form of metaphilosophy, it is a reflective or higher-order discipline that takes ordinary epistemology as its subject matter, which itself is a first-order or substantive discipline. Although there is a general agreement that metaepistemology reflects on epistemology in some sense, its exact definition is contested. Some sources define it narrowly as the epistemology of epistemology, including The Blackwell Dictionary of Western Philosophy which states that the role of metaepistemology is in comparing different epistemologies and analysing epistemic concepts. Others emphasise the role of metaepistemology in examining epistemology's goals, methods and criteria of adequacy. Metaepistemology is also sometimes characterised as the study of epistemic statements and judgements, including their semantic, ontological and pragmatic status, or as the study of epistemic facts and reasons.

Metaepistemology is a relatively modern term and probably originated at some point in the 20th century. Dominique Kuenzle identifies its first use as a 1959 article by Roderick Firth discussing the views of Roderick Chisholm on the ethics of belief. Early explorations into the concept of metaepistemology were undertaken by William Alston and Jonathan Dancy in the late 1970s and early 1980s, with Alston defining metaepistemology in 1978 as the study of "the conceptual and methodological foundations of [epistemology]." Kuenzle notes only a few other uses of the term prior to 2017, but Christos Kyriacou and Robin McKenna state that increasing interest in the field arose around the beginning of the 21st century due to a growing recognition that epistemology is a normative field like ethics. Like other terms in philosophy, metaepistemology pre-dates the name given to it; Kyriacou traces its study back to Plato's discussions of the value and source of knowledge, and ancient Greek scepticism.

The divisions between metaepistemology and the other branches of epistemology—as well as their connections with one another—are debated by metaepistemologists. For example, the Internet Encyclopedia of Philosophy divides epistemology into three branches, analogous to the three branches of ethics: metaepistemology, normative epistemology and applied epistemology. Richard Fumerton instead divides epistemology into metaepistemology and applied epistemology. According to Fumerton, the idea of a branch of normative epistemology is problematic because, in his view, epistemic normativity is inherently different in character to moral normativity.

Views about the relationship between metaepistemology and the other branches of epistemology fall into two groups: autonomy and interdependency. The autonomy view claims that metaepistemology is an entirely independent branch of epistemology—it neither depends on the other branches nor do they depend on it. According to this view, a person being a metaepistemological realist or anti-realist has no implications for what views they should accept in first-order epistemology. The interdependency view, on the other hand, claims that there are strong theoretical interdependencies between the branches and a first-order epistemological view may even directly follow from a metaepistemological one. Furthermore, according to the latter view, metaepistemology may have relevance to issues of practical importance like climate change scepticism.