<p>Impossible worlds have been introduced by many philosophers to model attitudes and propositions. However, the complexities associated with first-person attitudes are often overlooked by advocates of this approach. This paper aims to address this gap by developing a novel impossible-worlds semantics for first-person belief ascriptions and arguing for its superiority over the prominent centred-worlds approach. I begin by presenting a challenge that first-person beliefs pose to traditional possible-worlds semantics. Next, I formulate a semantics for first-person belief ascriptions derived from the centred-worlds approach, demonstrating how it addresses the issues faced by possible-worlds semantics. I then critique the centred-worlds semantics by highlighting several issues it faces. Finally, I propose an impossible-worlds semantics for first- person belief ascriptions and argue that it resolves the problems the centred-worlds semantics aims to address while avoiding the issues inherent in that semantics.</p>