RNA targeting represents a compelling strategy for addressing challenging therapeutic targets that are otherwise intractable through traditional protein targeting. Revolutionary approaches in RNA-focused small molecule libraries have successfully identified RNA-binding ligands but generally remain limited in diversity and impeded by a dearth of structural insight into RNA and RNA complexes. Cyclic peptides are potential structural mimics of evolutionary RNA-protein interacting motifs and can be massively diversified and selected via genetically encoded libraries, offering a complementary approach. This study introduces genetically encoded thioether cyclic peptide libraries constructed through mRNA display using a dibromoxylene linker and its fluorosulfonyl derivative that can covalently engage RNA nucleophiles. Using an optimized mRNA display workflow for RNA binders, we discovered high affinity, covalent and noncovalent binders for SNCA 5′ UTR IRE, the upstream iron-responsive element that post-transcriptionally regulates the expression of α-synuclein, an intrinsically disordered protein implicated in Parkinsonism and related neurodegenerative diseases. Notably, a stringent selection strategy employing “base-paired” target analog counterselection enhanced specificity by deenriching nonspecific electrostatic interactions mediated by polycationic residues. Further engineering hit peptides with an imidazole tag yielded selective RNA degraders in which covalent degraders showed noticeably improved potency from noncovalent counterparts. This work provides a prototype framework for evolution-driven, high-throughput, RNA-targeted drug discovery using cyclic peptides.