Comparing an entity’s financial indicators with those of similar organizations can provide a better understanding of its operational and financial health. This study describes the design and implementation of a prototype multilabel classification method to categorize nonprofit organizations (NPOs) using the textual content of their mission statements to enable beneficial comparisons. Positive unlabeled learning was used to improve the classification performance of partially labeled data. Naive Bayes, Gradient Boosting, Random Forest, and Support Vector Machine (SVM) algorithms were applied to determine the most effective method for classifying NPOs. The SVM model performed best in identifying “Housing and Shelter” organizations. The SVM classifier identified organizations that were not previously classified as “Housing and Shelter” but provided housing and shelter services as a part of their programs and activities. The new classification method can help donors, grant providers, and researchers to identify similar nonprofit organizations at the operational level.