Problem Statement: The interest in raising levels of achievement in math and science has led to a focus on investigating the factors that shape achievement in these subjects. Understanding how different learning styles might influence science achievement may guide educators in their efforts to raise achievement. This study is an attempt to examine primary school students' science performance on Trends in International Mathematics and Science Study (TIMSS) cognitive domains, based on their learning styles. Being aware of learning styles and their influence on different cognitive domains may provide educators with ideas for differentiating instruction and may help improve TIMSS achievement. Purpose of Study: This study examined the differences in 8th grade students' science scores in terms of the knowing, applying and reasoning domains of TIMSS, based on Kolb's learning styles and the relationship among learning modes and the TIMSS domain scores. Methods: A science test developed from the released TIMSS items measured 8th grade science achievement, and Kolb's Learning Style Inventory (LSI) determined the preferred learning styles. Relationships among students' learning mode and dimension scores and domain scores were examined through a bivariate correlation analysis. Differences in the total science scores of students in four types of Kolb's learning styles were examined through Analysis of Variance (ANOVA). Next, Multivariate Analysis of Variance (MANOVA) was conducted to examine the differences in knowing, applying, and reasoning domain scores based on learning styles. Findings and Results: The results showed that assimilating and converging learners were consistently more successful, while diverging learners were the least successful in all three cognitive domains of TIMSS science. The correlation between the Abstract Conceptualization-Concrete Experience dimension score and achievement increased as questions became more complex. Conclusions and Recommendations: It was concluded that students might need to utilize their abstract conceptualizing skills rather than their concrete experience skills in order to become successful in TIMSS assessments. It is crucial to assess students' learning styles in order to motivate educators to reflect on their teaching styles. There is strong empirical evidence that learners' performance has increased when teaching was arranged according to their learning preferences in higher education. Further emprical evidence is needed for whether learning-style- based instruction described by Kolb improves primary school students' achievement.