The agribusiness on the São Francisco Valley stands out in the national agribusiness by the mango and grape production, however it has been increasing the vulnerability to extreme weather events associated with thermal rises in Palmer mango crops. This cultivar has been producing stenospermocarp fruits with great frequency when the flowering and initial development of the fruits coincide with to higher temperatures. Thus, this study, capable of profiling the occurrence of high temperatures in the region may be important to accurately determine production risks associated with thermal extremes throughout the year, aimed to characterize the maximum air temperature on a monthly scale through descriptive statistical elements, as well as to identify the period with the greatest potential for inducing stonepermocarp in cv. Palmer in the Sub-Middle region of the São Francisco Valley. For graphical analysis of the data series, a histogram of frequency of the maximum observed temperatures was constructed each month, which was superimposed by the Normal probability density function. The following descriptive statistical elements were used: mean, mode and median, variance and standard deviation. Asymmetry measures were also used, such as the asymmetry coefficient and kurtosis measures such as the kurtosis coefficient. The use of the mean alone to feature the São Francisco Valley temparatures, can bring inconsistency and fragility, due to a considerable degree of uncertainty and inadequacy in the thermal characterization. The use of the descriptive statistical elements used in this study had been proved a viable tool for the thermal extremes features of the São Francisco Valley. November tends to be characterized as the month of greatest risk when the mango cv. Palmer, with observation of the extreme maximum temperatures, showing the highest monthly average, 34.1°C and frequent occurrence of temperatures of 37.5°C, therefore the month with the highest risk to ocurr stenospermocarp fruits.