<p>This paper systematically assesses the effectiveness and potential biases of interventions designed to enhance pro-environmental beliefs related to climate change (CC). Using the Cochrane risk-of-bias tool for randomized trials (RoB-2), we evaluate 12 randomized controlled trials (RCTs) retrieved from seven databases. Our findings revealed that there was no single intervention design or method that emerged as clearly the most effective. All but one trial had some concern or high risk of bias, suggesting a possible link between the predominance of interventions with small or no effect and the presence of biases. Our analysis highlights two prevalent biases: selection and instrumentation. These biases were found to undermine the accuracy of many studies, calling for future research to control these factors. Pre-test/post-test designs, along with establishing clear cut-off points for CC belief levels, could help mitigate selection bias. Additionally, the adaptation of psychometrically sound instruments for measuring CC beliefs is essential for reducing instrumentation bias. Four such instruments were identified, offering promising opportunities for more reliable measurement in future studies. We also observe a predominance of studies based in high-income countries, particularly the USA, which underscores the need for future RCTs to evaluate interventions in low- and middle-income populations. Overall, while challenges remain in designing effective interventions, the integration of these recommendations could improve the accuracy and generalizability of future climate change belief interventions. We discuss further research directions to guide the design of future RCTs and meta-analyses.</p>