I investigated two aspects of in vivo splicing that are poorly understood: spliceosome disassembly and recycling, and proofreading. To this end, I used the auxin-inducible degron (AID) to individually deplete several splicing factors in budding yeast and then I measured the effect on co-transcriptional spliceosome assembly through chromatin immunoprecipitation. In addition, using RNA next-generation sequencing, I measured the frequency of splicing errors following depletion or mutation of the fidelity factor, Prp22. I show that formation of the pre-spliceosome (the first stage of spliceosome assembly) is rapidly inhibited by global defects in late stages of spliceosome assembly. I demonstrate that this is due to the accumulation of arrested spliceosomes that sequester the splicing machinery and, as a result, causes a recycling defect. This suggests that spliceosomes that lack essential splicing factors are not always properly disassembled and recycled in vivo, and warns about potential systematic secondary effects when perturbing single components of the spliceosome. Secondly, I describe the development of a new version of the AID system for budding yeast, called the B-estradiol AID. To the best of my knowledge, an AID system for budding yeast that is fast-acting, tightly-controlled and gratuitous, was lacking until now. Lastly, I show that absence of Prp22 protein, which was previously proposed to play a role in splicing fidelity, correlates with more mistakes in 3’ss selection of many endogenous intron-containing transcripts in vivo. This provides indirect evidence to suggest that Prp22-dependent splicing proofreading is physiologically important. The data from this analysis will be useful in ongoing studies to try to identify common features that could improve our understanding of the mechanism of Prp22’s function in splicing proofreading.