The measurement of colour has been frequently used as an alternative to chemical analysis in a wide range of applications. Colorimetry and spectrophotometry are the reference techniques nowadays to measure colour. Practical problems arise, however, when granular, powdery, rough, curved or non-regular surfaces are processed. A standard digital camera, in combination with image analysis, can be used as an inexpensive, portable, non-destructive and fast analytical tool to measure colour but, surprisingly, this approach has not been fully exploited up to date. In this work, we have used this technique to measure tomato colour at different ripening stages. Spectrophotometric analysis of the extracts of the tomatoes allowed us to record the characteristic spectrum of each sample and correlate them with the chromaticity parameters by non-linear regression analysis of the whole data set. The resulting mathematical model showed a high power to discriminate between tomatoes, and was further used to produce a relative scale from zero to ten that enabled us to sort naturally matured tomatoes according to their ripening stage. In addition, the proposed model also allowed a fast estimation of the lycopene content in the samples, avoiding long and tedious extraction procedures, with a significant saving of time and reagents.