Sapodilla fruit is one of the tropical fruits that are widely cultivated in Indonesia, especially West Sumatra.  During post-harvest, especially during  the grading process, sapodilla fruit will be dropped into the basket at a certain height so  that it has the potential to cause mechanical damage that is not visible at the beginning of the drop. This study aims  to evaluate the potential use of e-nose  technology to detect mechanical damage to sapodilla fruit non-destructively at the beginning of dropping during  the grading process.   The method used is an experimental method by looking at changes in the VOCs content of sapodilla fruit through mics 5524 sensor readings.  The results showed that e-nose technology was  able to detect mechanical damage to  sapodilla fruit at the beginning of dropping with a sig value of  < 1% in ANOVA analysis, meaning that there was an influence of height at the beginning of dropping on changes in the VOCs content of sapodilla fruit. The results of grouping using k-means clusters show that  the height of the fall of 0 cm and 20 cm is included in the category of undamaged sapodilla fruit and is in cluster  1, while the height of the fall of 40 cm is included in the  category of damaged sapodilla fruit  and is in cluster  2.