Parametric effects in colour-difference evaluation were investigated using a CRT display controlled by a computer. A large number of visual data were obtained by a set of observers assessing colour-differences of sample pairs presented under 16 groups of parameters with grey-scale method. The results of preliminary processing and analyses indicated that the effects of sample frame and background colour on colour-difference perception were more obvious than other parameters considered in the experiments. The performance of BFD colour-difference formula for predicting colour-difference perception was the best and CIELAB the worst among four colour-difference formulae in comparison. Finally, two observing conditions were obtained for achieving the closest colour-difference perception of fluorescent colour on CRT and of surface colour on textile samples.