Targeted cancer therapy is a very promising concept; however, small molecule inhibitors only exist for a limited number of oncoproteins and resistance development limits their clinical use. Different engineered non-antibody scaffolds are being explored to develop potent protein-based inhibitors of oncoproteins, which can overcome these hurdles. The first goal of this thesis was to characterize monobodies, engineered binders based on the fibronectin type 3 domain, which were selected to bind the SH2 domains of 6 members of the Src kinase family that play a role in different types of cancer. I have shown that the monobodies bind specifically to their target in cells by performing tandem affinity purifications with subsequent mass spectrometry analysis. When expressed in T cells, the Lck-targeting monobodies ML1 and ML3 inhibit the phosphorylation of Zap70, a kinase which is activated downstream of T cell activation. Moreover, the fusion of ML1 or ML3 to VHL, the substrate receptor of an E3 ubiquitin ligase, resulted in the targeted degradation of Lck. The selected monobodies therefore have great potential to interfere with cancer cell signaling and to be used in therapeutic approaches. The development of protein-based inhibitors is currently limited to extracellular targets due to the inability of proteins to cross the cell membrane. Many approaches to achieve cellular delivery are hampered by non-selective cytotoxicity or endosomal entrapment of the protein cargo. The second aim of my thesis was to explore different methods to deliver functional monobodies into cancer cells. Firstly, I tested cell penetrating poly-disulfides (CPDs). CPD-monobody adducts showed toxic effects on the tested cell line, and mainly followed an endocytic uptake route leading to lysosomal degradation, when used at subtoxic concentrations. Secondly, bacterial toxins have naturally evolved to deliver their payload into the cytoplasm of host cells. Both Shiga-like toxins and Exotoxin A from Pseudomonas aeruginosa are taken up into the cytosol via a retrograde trafficking route, avoiding endosomal entrapment. To deliver monobodies to the cytosol, we have tested a combination of their non-toxic subunits: The B subunit of Shiga-like toxin (Stx2B), binding to the sphingolipid Gb3 in the cell membrane, and the translocation domain of Exotoxin A (TDP). We could show that recombinant Stx2B-TDP-Monobody fusion proteins are taken up in HeLa cells, which naturally express Gb3. Colocalization analyses with markers for endocytic compartments demonstrated that the major part of the monobodies escape early endosomes and are not degraded in lysosomes. A recombinant fusion protein of the Stx2B-TDP construct with the monobody AS25, which allosterically inhibits Bcr-Abl kinase, leads to apoptosis in Bcr-Abl-dependent chronic myeloid leukemia cells. Furthermore, the delivery of the VHL-ML3 fusion protein into Jurkat T cells results in decreased Lck levels. The delivery of both constructs is dependent on Gb3 on the cell surface as well as the Stx2B-TDP subunits. These results demonstrate that functional monobodies can be delivered to the cytosol and bind to their target protein in cells. This method could facilitate the development of protein-based inhibitors of intracellular proteins for cancer treatment.