One of Frege's most characteristic ideas is his conception of truth-values as objects. On his account (from 1891 onwards), concepts are functions that map objects onto one of the two truth-values, the True and the False. These two truth-values are also seen as objects, an implication of Frege's sharp distinction between objects and functions. Crucial to this account is his use of function-argument analysis, and in this paper I explore the relationship between this use and his introduction of truth-values as objects. In the first section I look at Frege's use of function-argument analysis in his first work, the Begriffsschrift, and stress the importance of the idea that such a use permits alternative analyses. In the second section I examine his early notion of conceptual content, and argue that there is a problem in understanding that notion once alternative analyses are allowed. In the third section I turn to his key 1891 paper, 'Function and Concept', where the idea of truth-values as objects first appears, and consider its motivation. In the concluding section I comment on Frege's general philosophical approach, which allowed objects to be readily 'analyzed out' in transforming one sentence into another.