Purpose – The purpose of this research is to find out how well students' critical thinking skills are in learning science, especially in mixed materials, by using the QODE (Questioning, Organizing, Doing, and Evaluating) and Discovery Learning models. Method – This research method applies quantitative methods. This study involved 49 students from class VIII A and VIII B as subjects with details of 25 students of class VIII A and 24 students of VIII B from SMP Negeri 2 Paron. The data analysis technique used prerequisite test analysis. The data analysis technique uses prerequisite analysis in the form of validity, reliability, normality, and homogeneity tests, while the hypothesis test uses an independent sample t-test. Findings – The QODE learning model is more effective in improving critical thinking skills compared to the Discovery Learning learning model on mixed materials at SMP Negeri 2 Paron with an average posttest value of 81.96> 78.00. The results of the hypothesis test of the posttest value proved that the tcount> ttable (1.677) while the tcount (QODE) was 2.138 and the tcount (Discovery Learning) was 2.128. This indicates a significant difference in the post-test results, which leads to the rejection of H0 and acceptance of H1 which indicates the effectiveness of the model. The result of the t-test value of QODE > Discovery Learning is 2.138 > 2.128 so it can be concluded that the QODE learning model is more effective than Discovery Learning in improving critical thinking in science lessons, especially mixed material for students of SMP Negeri 2 Paron. Research Implications – Students' critical thinking skills can be improved with QODE and Discovery Learning when learning science, especially mixed materials at SMP Negeri 2 Paron where the QODE model is more effective than Discovery Learning.