Consuming alcoholic beverages improperly or excessively has a very bad impact on the health of the human body, it can even cause accidents, fights, theft, murder and the risk of death. This research produces a tool that can detect alcohol content in liquid herbal using the MQ-3 sensor, DHT22 sensor, data logger to store test results data and the test results are grouped by means of manual calculations with Microsoft Excel and Matlab software simulations. The type of liquid herbal used is liquid herbal for rheumatic pain, gout, dizziness, colds and coughs. In each liquid herbal test, it takes 10 minutes/liquid herbal medicine and is repeated 6 times for each liquid herbal test. Based on the test that the higher the temperature affects the increase in the proportion of the alcohol content of the liquid herbal and the minimum temperature obtained is 30.8°C to 32.9°C. Microsoft excel calculations and matlab simulation software, clusters are divided into 3 according to the type of alcohol content group and the iteration stops at the 12th iteration with the objective function value = 0.00074 < error = 0.001. From the results using Matlab Software, 8 data were obtained in group A with an average alcohol content 9,9%, 9 data in group B with an average alcohol content of 9,05% and 13 data in the content group C with an average alcohol content of 10,73%.