Pre-mRNA splicing is a eukaryotic phenomenon enabling multiple protein isoforms to be encoded by the same gene; the protein sequence can be altered by the use of alternative splice sites. Using RNA-Seq data I explored the full complement of introns excised from pre-mRNAs in the model organism Neurospora crassa to investigate splicing of introns with noncanonical splice site sequences. I report the splice site pairs observed in my data and provide evidence that AT dinucleotides can function as splice acceptors in N. crassa. This is the first time that AT dinucleotides have been reported as splice acceptors outside of rare occurrences in the human transcriptome. I present the serendipitous discovery of the first reported N. crassa twintron and show that N. crassa introns can be at least twice as long (~3.6 kb) as the NC12 genome annotation suggests. These results highlight the limitations imposed when transcriptome analysis considers only the most widely accepted features of introns valid. Light is an important stimulus for N. crassa biology, including the onset of asexual development and the synthesis of photoprotective pigments. To investigate the contribution of alternative splicing to the N. crassa light response, I generated a measure of intron retention, the most common mode of alternative splicing in lower eukaryotes. Candidate introns with differential read coverage between dark-grown and light-pulsed cultures reveal intronic read distributions consistent with alternative transcription start sites, rather than intron retention. Thus, it is unlikely that light-regulated splicing is a phenomenon acting in N. crassa. Introns exist in a branched lariat conformation after splicing, but are rapidly debranched and degraded. I created a Neurospora crassa knockout strain lacking RNA lariat debranching activity, which enables enrichment of excised introns and will facilitate future efforts to understand pre-mRNA splicing in this organism.