Skin can be considered the largest organ in the human body. Healthy skin is not only good for the body, but alsoenhances the appearance. Good skin care is essential at any age. In the first few decades of life, the skin has aconsiderable supply of elastin and collagen, but it will gradually decrease. In addition, daily lifestyle can also directlyaffect the appearance of human skin. The purpose of the research is to develop a model that classifies facial skin typesbased on physiological data using random forest algorithm and measure the results of accuracy, precision, and recall.This research uses Rapidminer tools and four facial skin types namely dry, combination, normal, and oily. The resultsof random forest research obtained accuracy results of 93.25%. dry precision 98.02%, combination precision 92.94%,normal precision 93.46%, and oily precision 88.79%. While dry recall 99%, combination recall 79%, normal recall100%, and oily recall 95%. The findings of this research can help create a skincare recommendation system that ismore suited to the needs of each individual.