The current understanding of cracking/splitting in grape and other fruit, is based on the assumption that flesh cell turgor pressure pushes outward against the skin, causing the skin to break, similarly to the failure of a thin shell. Based on this understanding, larger berries and berries with thinner skin should be more cracking susceptible, but research to date has not found any consistent relation between cracking susceptibility and these, or other such berry structural features. In fact, previous research in our lab has shown that the intact skin of the cracking susceptible variety ‘Flame seedless’ is capable of stretching much more without breaking (about 7.5% strain) when it is isolated from the flesh, compared to 1-2% strain associated with cracking in intact fruit. Based on multiple lines of evidence (partial soaking, soaking of sectioned berries or berries partially coated with Vaseline), we have also found that the berry skin does not need to be entire to crack, and hence does not crack as would a uniformly stressed thin shell. In this study, time-lapse microscopy and computer image analysis was used to measure the development of skin strain for grapes submerged in water. We found that skin strain was locally uneven (non-uniform) at the future site of crack initiation, and that this occurred substantially (20 minutes) prior to crack appearance. In contrast, strain was uniform in nearby non-cracked areas of fruit that cracked, as well as areas of fruit that did not crack. Due to technical limitations of the image analysis program however, these results were limited to measurements of 1D (linear) strain, and only applied to specific sites on the fruit skin. Further progress required the development of an image analysis program capable of measuring 2D (area) strain over a more complete surface of the fruit. Time-lapse images of over 40 fruit, including Flame seedless from 4 locations and DOV berries, was collected in 2015, but due to changes in personnel, the 2D image analysis program was not completed until the end of the 2015/16 academic year. We are currently re-analyzing the 2014 time-lapse data, which thus far has confirmed that 2D non-uniformity is also localized to the area of crack initiation and growth. These results indicate that the current thin shell model of fruit cracking is inadequate to explain the process of crack initiation, and further suggest that cracking susceptibility may be caused by factors such as localized pressures on the skin, rather than related to the strength of the skin itself. The 2D approach however contains substantial additional information (such as the direction of the strains), which may be critical in understanding the cracking process itself, and we are continuing analysis of the 2014 data to determine the best approach to apply to the 2015 data. This analysis should be complete by fall of 2017.