Epistemically suspect beliefs, such as endorsement of conspiracy theories or pseudoscientific claims are widespread even among highly educated individuals. The phenomenon of conspiratorial thinking is not new, yet the COVID-19 pandemic, causing a global health crisis of an unprecedented scale, facilitated the emergence and rapid spread of some rather radical health related pseudoscientific fallacies. Numerous correlates of the tendency to endorse conspiracy theories have already been addressed. However, many of them are not subject to an intervention. Here, we have tested a model that includes predictors ranging across stable characteristics such as demographics (gender, age, education, size of the place of residence), less stable general traits such as conservatism and overconfidence in one’s own reasoning abilities, to relatively changeable worldviews such as trust in science. A hierarchical regression analysis (N=859 participants) showed that included predictors explained a total of 46% of the variance of believing in COVID-19 conspiracy theories, with only gender, overconfidence and trust in science yielding significance. Trust in science was the strongest predictor, implying that campaigns aimed at enhancing public trust in both science as a process, and scientists as individuals conducting it, might contribute to the reduction in susceptibility to pseudoscientific claims. Furthermore, overconfidence in one’s own reasoning abilities was negatively correlated with an objective measure of reasoning (syllogisms test), and positively with the endorsement of conspiracy theories, indicating that so-called Dunning-Kruger effect plays a role in pseudoscientific conspiratorial thinking regarding COVID-19.