Choosing a major is a crucial decision for prospective students entering higher education. An inappropriate choice may lead to low learning motivation, poor academic performance, and career mismatches. This study aims to develop a multiclass majors recommendation system based on machine learning using the Random Forest algorithm. The dataset consists of 855 student and alumni records from 10 majors at Dharma Andalas University (UNIDHA), including academic attributes (subject grades, GPA, entrance test results) and non-academic attributes (gender, high school major, interest, and alumni career field). The model was trained using an 80:20 train-test ratio and evaluated using accuracy, precision, recall, F1-score, and macro-average AUC. The results show that the Random Forest outperforms Decision Tree, K-Nearest Neighbor, and Naive Bayes, achieving an accuracy of 0.920 and AUC of 0.972. These findings demonstrate that ensemble-based algorithms are highly effective for multiclass recommendation problems and can serve as a foundation for academic and career guidance systems in universities.