In order to investigate the firmness of tomato slices, two experiments were performed. In the first one, Monte Carlo simulation was used to study the variation in firmness within and between slices. Adding more slices and more measurements per slice reduced the SD, but in general, the efficiency of adding more slices was higher. In the second experiment, the firmness of tomato slices was measured by puncture test during storage, using one of three flat-tipped cylindrical probes (3.5-, 2.5- and 1.5-mm diameter) in two directions, along or perpendicular to the main axis of the fruit. Changes in firmness were studied by nonlinear regression analysis. The same model could be applied to all combinations of probe size and direction with the same correction for shear and compression. It suggests that shear and compression forces decay with storage time according to the same mechanism, irrespective of the measurement direction.