Along with the increasing world of the pharmaceutical industry, product information becomes input for companies. One of them is drug sales information and drug supply information. Information on how much drug supply is very important because this is related to how many sales will occur or the target market will be achieved in a certain period of time. The sales pattern is an activity to find out how consumers buy products with reference to maintaining the stability of stocks in increasing sales. The author helps the pharmacists of Srikandi Pharmacy to look for forecasting results in the future period by using the Supervised Learning Algorithm to perform calculations using the Support Vector Machine method as a method for calculating classification and prediction processes with high-dimensional space features. There are 344 data which are divided into two categories namely High and Low which can be seen from the pattern of sales. Before looking for prediction results here, we determine the best kernel to look for a high accuracy value to get optimal calculation results. Calculation test results using SVM get an accuracy of 99.2% with Polynomial kernels and C = 1.00, λ = 0.10, g = 000.1, d = 3.0.