This article looks at the known foundational results on spatial models of elections. The issues of equilibrium existence, the characterization of equilibria (in terms of their social welfare properties), and the distance between equilibrium policy and positions of the candidates are examined. It then discusses the results of the case where candidates are able to give precise predictions of voters' behaviour precisely; the article also introduces the ‘Downsian model’. The article looks at two models of probabilistic voting, before finally moving on to consider the most common objective functions that are used to model the electoral incentives of different types of candidates.