Displays such as mobile phones are viewed under surround conditions that vary from dark night to bright sunlight. The overall goal of this study is to test the appropriateness of CIECAM02 for mobile displays, and modify it according to any insufficiency found. Firstly, the testing of CIECAM02 is described using the visual data from 2‐inch display of three achromatic backgrounds (gray, black, and white), and three surround conditions (dark, dim, and average). Secondly, CIECAM02 was tested under four surround conditions (dark, dim, average, and bright), and three displays (2‐, 4‐, and 7‐inch), with only the gray background used, to focus more on the surround conditions. Those twelve experimental phases were used to refine CIECAM02, considering the surround factors. The surround parameters (c, F, and NC) were optimized, and Ncb was modified from the CIECAM02; the modified model is named MobileCAM‐v1. Colourfulness prediction by MobileCAM‐v1 has been improved, especially for a black background under average surround condition. A further refined version of CIECAM02, MobileCAM‐v2, was developed, for mobile displays viewed under different surround conditions. A set of equations based on surrounding conditions was first derived, to be able to accurately define surround parameters. The MobileCAM‐v2 model gave the largest improvement in brightness, followed by lightness, and with colourfulness the smallest. The improvement is significant for bright surround conditions, improving the performance of CIECAM02 in predicting the visual results. © 2014 Wiley Periodicals, Inc. Col Res Appl, 40, 114–124, 2015