The convergence of Non-Orthogonal Multiple Access (NOMA) and Cognitive Radio (CR) with Simultaneous Wireless Information and Power Transfer (SWIPT) offers a transformative approach to spectrum and energy efficiency. This paper analyzes CR–NOMA with SWIPT-enabled energy harvesting (EH), focusing on outage probability (OP) and throughput. We derive explicit models under Rayleigh fading with interference-temperature constraints and validate by simulations. Results show that proper power allocation and time-switching ratios enhance user fairness while respecting primary-network interference. These insights guide robust and energyefficient designs for next-generation systems.