A psychophysical experiment for describing image quality of a 2-inch QVGA mobile liquid crystal display (LCD) was carried out using a category judgment method. Five natural test images were rendered in terms of 8 physical parameters: peakwhite luminance, resolution, bit depth, correlated colour temperature (CCT), lightness with linear and non-linear alterations, chroma and hue. Ten observers rated each of the rendered images using 9 categories (1 to 9), according to 7 perceptual attributes: naturalness, clearness, sharpness, contrast, colourfulness, quality and preference. In total, 18900 judgments were made. The whole set of psychophysical data were used to build an empirical image quality model and, through a stepwise regression method1, it was found that naturalness and sharpness are the most important attributes quantifying image quality.