In the research that is applied to examine the impact of science education activities based on STEM approach on preschool term children’s science process skills, used pre-test post-test control group semi-examinational model. 38 children and their teachers, being 20 experiments and 18 control groups continuing to kindergartens of elementary schools in middle income region of Çankaya District of Ankara Province, are included in the research. In order to prevent and control groups affected from each other, children are selected from different schools. To collect data Science Process Skills Scale for 60-72 Months Old Children and Demographic Information Form are used. One day per a week for 12 weeks, science education activities based on STEM approach are applied to experiment group children by researcher. In the end of the research after data are collected, 3 sessions are made to control group children with ethical reasons. To determine the difference between pre-test and post-tests applied to experiment and control groups Independent Samples T-Test is applied, while Paired Samples T-Test is applied in order to determine the difference between pre-test and posttest of experiment and control groups. It is determined that there is no meaningful difference in pre-test scores of children in experiment and control groups in terms of science process skills according to findings obtained as a result of the analysis data, while there is a statistically meaningful difference in advantage of experiment group between posttest scores. Results show that science education activities based on STEM approach provide effective contribution to the development of science process skills of children who in the experiment group. As a result of the research, it is possible to said that science education activities based on STEM approach have a positive impact on especially children’s performance about predicting, linking conclusions to data, sharing findings with others and observing.