Iris colour can provide an enormous amount of information about an individual. In addition to changes with pathological conditions, the colour of the iris can be a particularly useful indicator of how well a person will respond to a topically applied ocular drug. Until recently, classification of iris colour has been subjective, ranging from a basic description (‘light’ and ‘dark’) to more detailed grading systems, such as a comparison with preset photographic standards. However, variability within observers and differences in the interpretation between observers can influence the results. Objective techniques, in this respect, possess several advantages. They are able to detect differences in colour that subjective techniques are incapable of and they provide continuous data rather than discrete categories, thus improving the accuracy of drug response predictions. This study assessed iris colour by objective means. Slit-lamp photographs of various coloured irides were taken under standardised conditions. The slides were then scanned into a computer and the colour analysed using a calibrated software package. To establish the optimum colour parameter to be used for predictions of drug response, several parameters were calculated and compared with the subject response to 1% tropicamide (maximum change in pupil size, time to maximum change and total duration of effect). Many parameters had strong correlations with drug response, but the parameters ‘z’, ‘b’ (the proportion of blue in the image) and ‘y’ (the proportion of yellow in the image) were found to exhibit the highest correlations. They also showed better correlations with drug response than did a current iris colour grading system.