Many targeted cancer therapies fail to improve the overall survival of patients. Limitations involve low drug selectivity and the rapid development of resistance. Moreover, the majority of oncogenic drivers presently remain considered as undruggable as these proteins often lack well-defined binding pockets able to fit small molecular antagonists. The need for novel strategies to expand the number of druggable targets thus appears crucial. STAT3 is a transcription factor that is constitutively active in a majority of solid and hematological tumors. It is considered a central target for cancer therapies, as it induces the transcription of genes essential for tumor proliferation, including anti-apoptotic genes, cell cycle regulators and angiogenic factors. Therefore, the inhibition of STAT3 represents a promising therapeutic strategy. However, the targeting of STAT3 still remains a formidable challenge due to the sparse number of chemical probes able to bind to non-enzymes. The engineering of protein binders could overcome many of the challenges in developing STAT3 antagonists. While most common approaches to preclude STAT3 activation consist in either inhibiting its phosphorylation by upstream tyrosine kinases with clinically-approved drugs, or in preventing the SH2 domain-dependent STAT3 dimerization using pre-clinical small molecule probes, the targeting of other STAT3 domains, such as its Coiled-coil or N-terminal domains, has not been thoroughly investigated. Our lab uses small engineered antibody mimics derived from a fibronectin type 3 scaffold (termed monobodies) capable of high affinity and selective binding. In addition, their small size (~10 kDa) and lack of disulfide bonds makes them promising candidates for intracellular antagonistic use. In this work, we selected several monobodies using phage and yeast display that bind to previously untargeted domains of STAT3 with low nanomolar affinities. The monobodies MS3-6 and MS3-N3, binding to the Coiled-coil and N-terminal domain of STAT3 respectively, showed high selectivity to STAT3 as compared to other STAT family members and other unrelated proteins. Additionally, the monobodies strongly inhibited the transcriptional activity of STAT3 in luciferase reporter assays, reduced mRNA levels of STAT3 downstream genes in human lung cancer cells, decreased STAT3 phosphorylation levels upon IL-22 stimulation and interfered with STAT3 nuclear translocation. Notably, the precise blockade of these previously untargeted key domains using monobodies provide new insights into the STAT3 signaling and expand the current strategies to preclude its activity. Altogether, we have developed the first selective monobody inhibitors against a transcription factor implicated in cancer and inflammatory diseases, and used them as biochemical tools to further characterize a poorly understood alternative IL-22R signaling axis involved in STAT3-driven human disorders such as colitis and psoriasis.