The research aims to compare the advanced class (experimental class) and traditional class with the traditional teaching method. This research is concerned with STEM activity in parachute design. The experiment class applies research-based learning methods integrated with Cloud Classroom (CCR), and the Traditional class is taught using conventional teacher talk or exp lanation. This research used triangulation methods , quantitative, and qualitative methods. The data analysis of the post-test indicated a value of sig (2-tailed) 0.00 where p≤0.05, wh ich implied the student achievement on combinatorial thinking skills criteria in the experimental class was better than that in the control class. The combinatorial thinking skills in experimental class after imp lementation of RBL was that 16% student were categorized in a low level of combinatorial thinking skills, 39% in med iu m level and 45% of the student was at the high level of comb inatorial thinking skills. The activity of students in the experimental class reached 35 % as the highest activity for very active activity, 27 % active, hesitate 27 %, inactive 6 %, and very inactive 5 %. Based on this result, the imp lementation of research-based learning integrated with CCR is proven effective in imp roving student Comb inatorial Thinking skills in STEM activity for parachute design.