Glioblastoma (GBM) is a cancer of the brain that has a high mortality and morbidity rate. In the current study, we aimed to find potential small molecules (50-1500 Dalton) with antiproliferative effects on the human U-87 GBM cell line using a machine learning (ML) based screening tool. A list of 519 small molecules that have been tested regarding their antiproliferative effect on the U-87 cell line in MTT assay after 72 hours was obtained from the ChEMBL database. Then feature extraction was conducted using Molecular ACCess Systems keys (MACCS) and Morgan fingerprints. Then, molecules in the lower 20 percentiles regarding IC50 were categorized as high anti-U-87 activity molecules, and other molecules were categorized as low anti-U-87 activity. A random forest model was trained on the data to build a screening algorithm. The list of 3421 clinically approved small molecules was obtained from the ChEMBL database. Among them, 8 were common with the training dataset so they were deleted. Then the remaining molecules were screened using the generated screening tool. Finally, 71 molecules from the approved dataset were classified as having an anti-U-87 effect, and based on our search some of them are approved drugs for GBM, some of them are approved drugs for other malignancies and some haven't been investigated regarding anti-neoplasm activity. Novel molecules that are found to have potential anti-U-87 effects in this study should be investigated in future in-vitro and in-vivo studies. Current methods and results could be used to design further studies exploring novel anti-neoplasm drugs.