A few years after his adoption of both a syntactical method and a principle of tolerance in the early thirties, Carnap turned to semantics when he learned about Tarski’s work on the definition of a truth predicate. How significant is this semantical turn? Carnap scholars have so much emphasized that The Logical Syntax anticipates Tarskian semantics that they tended to minimize the importance of Carnap’s adoption of a semantical approach. As a consequence, his semantical turn has not always been given the importance it deserves. Its meaning, scope and consequences have also often been misunderstood. This paper contributes to a re-evaluation of Tarski’s influence on Carnap in view of the fact that Carnap is far from having just followed Tarski’s way. We examine some specificities of Carnap’s approach of semantics. We also discuss what is left of the syntactic method after the adoption of semantics and what the relations between syntax and semantics become from the late thirties on. The following topics are given specific attention: languages, formal systems, and calculi; truth, L-truth, and L-validity; L-states, L-ranges, and state descriptions. We also analyze the impact of semantics on the principle of tolerance.