Quantum computing faces significant challenges due to quantum noise and gate errors, particularly in noisy intermediate-scale quantum [9] devices. Traditional error mitigation methods often fall short of achieving high circuit fidelity due to inherent system imperfections. This paper introduces Quantum-Classical Evolutionary Optimization [1], a novel hybrid framework that integrates quantum circuit mutation strategies with classical optimization techniques. QCEO adapts quantum circuits in real-time by treating quantum gates as evolutionary genes, iteratively optimizing them for improved noise resistance and fidelity. In experiments on both simulated environments and IBM Quantum hardware, QCEO demonstrates a 10-15% improvement in fidelity compared to traditional methods such as readout error correction and zero-noise extrapolation. Beyond fidelity improvement, QCEO’s ability to dynamically adapt to different noise profiles positions it as a versatile tool that can be integrated with emerging quantum algorithms like Variational Quantum Algorithms [4] and Quantum Approximate Optimization Algorithms [10]. This framework opens up new possibilities for advancing quantum computing in fields ranging from cryptography to quantum machine learning by ensuring higher performance in noisy quantum systems.