ABSTRACT. As cloud computing is becoming more important, there is a growing urgency for the development of more efficient and secure encryption frameworks that are cost-effective, scalable, and secure. Unfortunately, traditional encryption such as RSA and fully homomorphic encryption (FHE) is highly computationally overhead and not feasible for large-scale cloud applications. To address these limitations, a residue number system-based partial homomorphic encryption (RNS-PHE) framework takes advantage of carry-free modular arithmetic and the parallel computing ability of RNS when computing encryption and decryption. RNS-PHE was used to perform the performance evaluation, and showed a significant improvement over the traditional encryption techniques. Encryption times of the framework were five times faster than RSA and ten times faster in decryption than FHE, while consuming less memory and scaling linearly in the context of a multi-tenant cloud. Brute force attacks and key compromise assessments confirmed RNS-PHE was an acceptable alternative for secure cloud computing. RNS-PHE offers a computationally efficient and secure encryption approach to preliminarily gaining cloud storage, secure computation, and encrypted data analysis in modern cloud environments.