Abstract: Functional non-coding RNAs play important biological roles within the cell and often have specific structures for the task at hand. RNA folding studies aim at understanding how the RNA achieves its final conformation. Previous studies have largely lacked a structural framework to describe the folding process. The present studies begin to provide such a framework. I present the first nucleotide-level structural model of an RNA folding intermediate using a barrage of biophysical and biochemical techniques. The equilibrium folding intermediate (I_eq) is from the specificity (S) domain of the Bacillus subtilis RNase P RNA. The intermediate lacks the native core and several long-range interactions among peripheral regions. To further probe I_eq, I employ electron cryomicroscopy (cryo-EM) together with single particle reconstruction. For the B. subtilis S-domain I_eq, the single-particle reconstruction has remarkable similarity to the structural models I created. Furthermore, the reconstructed image is used to improve our structural model. This cryo-EM study is the first of its kind on molecules ∼50kDa. To understand the structural basis of RNA stability, I investigate equilibrium folding of the S-domain from a mesophilic (E. coli) and a thermophilic (T. thermophilus) bacterium. Phylogenetic comparison reveals that 12 nucleotide differences that contribute to higher stability. The crystal structure of T. thermophilus reveals that these residues participate in extensive networks of hydrogen bonding, stacking and metal ion coordination throughout the molecule. A mesophilic S-domain mutant containing these 12 nucleotides has the same stability and folding cooperativity as the thermophilic S-domain. Finally, I investigate the kinetic folding pathway of the B. subtilis S-domain, which involves the formation of two transient intermediates, I_1k and I_2k, whose interconversion is the rate-limiting step. The rate-limiting step is insensitive to NaCl, urea, and D₂O but is sensitive to the divalent metal ion identity used in folding. The rate-limiting step is a small-scale event, perhaps involving structure formation around a pre-bound metal ion and/or a rearrangement of some base pairs. Lastly, I present two initial models of the first kinetic intermediate, I_1k.