The fate of RNA molecules is dependent on their interactions with RNA-binding proteins (RBPs), which regulate splicing, translation, decay, epitranscriptomic changes and transport. Increased understanding of RNA biology advocates that many disorders are caused by RBP-related defects, including cancer and neurodegeneration. This highlights the necessity for a comprehensive and systematic snapshot of RBP-RNA interactions in different cells, tissues, and disease-contexts. We and others have established methodologies allowing the global identification of RBPs from cell culture experiments. Here, we introduce the brain Peptide Cross-Linking and Affinity Purification (brain-pCLAP) methodology, an improvement of our previously published mass-spectrometry-based strategy, enabling global identification of RBPs and their active RNA-binding domains (RBDs) directly from tissue. Applied to the mouse brain, we identified 526 RBPs and their active RBDs. 86 were novel, including many synaptic vesicles proteins. Selected candidates were validated by Cross-Linking Immuno-Precipitation (CLIP). We next applied brain-pCLAP to a mouse model of Huntington’s disease (HD), for which splicing defects have been observed. We identified a single RBD, belonging to the alternative splicing factor RBM5, for which binding was changed in HD. Similar amounts of RBM5 proteins between normal and HD brains precluded differential expression as an explanation for the change in binding. However, global mass-spectrometry identification of RBM5-protein interactors by co-immunoprecipitation showed an imbalanced partnership between RBM5 and other splicing factors. Using CLIP, we identified the RNA targets of RBM5, many of which encode proteins involved in neurodegeneration, and confirmed its differential binding on these in HD. RNA-Seq analysis further showed that the RBM5 targets are differentially spliced in the diseased brain. RBM5 knock-down and overexpression in human neurons confirmed the conserved role of RBM5 in controlling splicing of these RNAs. Altogether, our data support the effectiveness of our approach in vivo and propose a mechanistic explanation for the aberrant splicing observed in HD.