The purpose of this research is to find out the science process skills of students through discovery learning. The process of science is essential for students to have because it is the process of forming science. Science process skills can accustom students to learning through scientific work processes and systematic work. The research method uses a quasi-experiment with one group pretest-posttest design. This study does not use a comparison class, because it already uses a preliminary test so that the magnitude of the effect or effect of using discovery learning can be known with certainty. In this study, research subjects were first given a pre-test (pretest) to find out the extent of the students' initial abilities before being given physics learning by using discovery learning. After the initial test is given, then the student is given treatment, namely learning physics by using discovery learning. After completing physics learning with discovery learning, then all students are given a final test (posttest) to determine the extent of the effect of physics learning by using discovery learning on students' science process skills. The instrument used in this study was a science process skills test on the material elasticity and Hooke law, in the form of 20 multiple-choice tests and tested on 30 students in Prisma City High School. Each question evaluates aspects of conclusions, observations, identification, predictions, and interpretation of data. The results of the analysis show an increase in observational indicators from 39.95 to 66.62, concluding from 53.86 to 69.4, identifying from 36.6 to 73.3, predicting from 39.43 to 68.31 and interpreting from 73.3 to 98.3. Based on the results of the analysis, the learning of physics-based on natural phenomena through discovery learning models can improve students' science process skills.