One of the oldest debates in psychological research into politicized science such as nanotechnology, vaccination, or climate change centers around the role of knowledge. Does increased knowledge of the science affect beliefs about it? While research has traditionally focused on the role of object-level knowledge, here we highlight the importance of meta-knowledge: How much people believe they know about the science. Specifically, we demonstrate the importance of meta-knowledge (measured as confidence in knowledge) for explaining beliefs about science with one of the most contested examples: climate change. For a national Germany sample (N = 509), frequentist and Bayesian analyses demonstrated that climate change meta-knowledge was predictive of climate change beliefs, above and beyond object-level climate change knowledge. These results held for both the belief that climate change is risky, and the belief that climate change is anthropogenic, and when controlling for political attitude, and demographic variables. Furthermore, for a second national German sample (N = 588), confidence in climate change knowledge was a stronger predictor of climate change beliefs compared to confidence in other-domain (biological and physical) science knowledge, suggesting that outside of the respective domain, metacognitive confidence did not explain beliefs. These results highlight the relevance of domain-specific metacognition for explaining beliefs about the contested science of climate change. By demonstrating the relevance of metacognitive, rather than solely object-level thought, these results add to our understanding of the cognitive mechanisms involved in the formation of beliefs about politicized science.