Cancer incidence is increasing and novel anticancer drugs with new mechanisms of action are essential for future chemotherapeutic treatment. Natural products have historically played an important role in the development of anti-cancer drugs and have potential to do so also in the future. In this thesis two classes of natural products are identified as possible drug lead candidates, and the mechanisms of their action are elucidated. Initially, in a screening of a compound library for cytotoxic effects in colon cancer cells, natural products with potent activity were identified. Based on their potency, and on previously reported activities in cancer cells, two main groups of compounds, cardiac glycosides (CGs) and gambogic acid (GA) analogues, were selected for further in-depth studies. The concentration-dependent cytotoxicity was confirmed in cell lines of different origin. Cardiac glycosides were mainly evaluated for their activity in colon cancer cells and in leukemic cells, whereas the GA analogues were studied using a resistance-based panel of ten human cancer cell lines. Using activity profiles and the ChemGPS-NP model, the compounds were compared, structurally and mechanistically, to standard chemotherapeutic drugs. The results from these analyses suggested that the CGs and the GA analogues act by mechanisms different from those of antimetabolites, alkylating agents, topoisomerase I and II inhibitors, or tubulin-active agents. By analysis of drug-induced gene expression, one GA analogue, dihydro GA, was identified as a possible inhibitor of the ubiquitin-proteasome system (UPS), and the CGs showed similarities to protein synthesis inhibitors. Starting from these hypotheses, we further investigated the mechanisms of actions on a molecular level. The results showed that GA and dihydro GA act as inhibitors of the 20S proteasome chymotrypsin activity, leading to accumulation of ubiquitinated proteins. The CGs were confirmed to inhibit protein synthesis in colon cancer cell lines. However, interestingly, in leukemia cell lines, it seemed that the CGs act through a different, yet unexplored, mechanism of action. The leukemic cells (pre-B and T-ALL) were particularly susceptible to the cytotoxic effects of CGs, including at concentrations that may be achievable in the clinic.