SignaltoNoise Ratio (SNR) is a key parameter for low light level imaging systems. It determines the performance and image quality of an imaging system. A SNR calculation model of low light level imaging systems based on CMOS imaging sensors is given. The relationship between the SNR and the signal and noise of a system is simulated. A SNR test environment is set up and SNR test experiment is completed. The experimental results show that the theoretical value is consistent with the measured value. Finally, the parameters of a given system are optimized according to the SNR analysis results. The calculation results show that the optimized system improves its SNR up to 4.5 under 1 mlux. This study of SNR provides the theoretical basis for the overall design and optimization of low light level imaging system based on CMOS image sensors.