This chapter turns to thoughts. Given they are the logician’s objects of investigation, Frege believes they must be purified for purpose. This endeavour requires us to know when two sentences are equipollent, i.e. when two sentences express the same thought. Frege believes that equipollence can be determined according to a criterion. I discuss the equipollence criterion in the early sections of this chapter, along with the insights it provides into Frege’s conception of thoughts. I then summarise Frege’s other claims regarding thoughts that appears in the published Logical Investigations series, before continuing onto his unpublished entry, Logical Generality, and examining the way Frege thinks a thought’s structure can be analysed. In so-doing I seek to do two things. First, to explain the connection Frege sees between natural language and the thoughts it expresses, including the way language offers the logician a tool for analysis. And second, to explain the relation Frege sees between the structure of a thought and the sentence that expresses it. After that, I defend the consistency of Frege’s claims regarding equipollence, the compositional nature of thoughts, and the analysis provided by expressing sentences against several objections raised.