The transformation of device responses to CIE XYZ tristimulus values is important at an early stage in color management and this transformation is usually called color correction. It is well known that the accuracy of color correction is greatly influenced by its spectral sensitivities and noise present in the device. However, the evaluation of the influence of noise on the accuracy of color correction has not been adequately addressed.This paper proposes the noise influence factor (NIF) to predict the influence of noise on the accuracy of color correction based on a new simple formula for a colorimetric evaluation of the device in the presence of noise. Computer simulations were performed to confirm the relationships between the NIF and the increased color correction errors by the presence of noise for each combination of sensor sets, color charts, recording and viewing illuminants and noise levels. It is shown that the proposed NIF provides a good measure to predict the influence for a variety of color sensor sets, color charts and illuminants. This predictor may be useful to design the optimal sensor sets that are robust to noise.