This study was conducted to examine the achievements of students in the concepts of light and shadow measured in different assessment formats according to the learning style and gender. In this study, correlational survey model was used. The sample consisted of 10th grade (16 years) high school students (n=815) from different types of six high schools (public general and vocational) in Turkey who were selected by using stratified random sampling method. The students' achievement was determined using three different assessment formats, which contains open-ended test, multiple-choice test and structured communication grid test. The results show that there were significant differences between the students' achievement in light and shadow concepts when assessed by different assessment formats. While the achievement of boys was statistically higher in open-ended tests, girls were more successful in the structured communication grid tests. In addition, the test scores obtained by students in different test formats vary significantly according to their learning styles. For example, the mean scores of converger and accommodator-style students in open-ended tests were significantly different from the diverger-style students.