The effects of cuticular crack surface area and inoculum density on the infection of nectarine fruits by conidia of Monilinia laxa were studied using artificial inoculations with conidial suspensions and dry airborne conidia during the 2004 and 2005 seasons, respectively. Additionally, the effect of ambient humidity on fruit infection was evaluated in the 2005 experiment. An exploratory analysis indicated that (i) ambient humidity did not significantly explain the observed variability of data, but that (ii) the incidence of fruit infection increased both with increasing inoculum density and increasing surface area of cuticular cracks. The product of these two variables represented the inoculum dose in the cracks, and was used as a predictor of fruit infection in the model. Natural infection in the orchard was observed to increase throughout the season in both 2004 and 2005. The relationship between the probability of fruit infection by M. laxa and the artificially inoculated dose in the cuticular cracks was well described by a logistic regression model once natural inoculum density was taken into account (pseudo R2 = 65%). This function could be helpful for estimating the risk of fruit infection at harvest based on fruit size and natural inoculum density