Pre-mRNA splicing requires assembly of spliceosome that consists of hundreds of factors forming various dynamic complexes with or without small nuclear RNAs (snRNAs). Systematic identification of the splicing factors remains a significant challenge especially in vivo. In our genetic screening for the factors required for division asynchrony during Caenorhabditis elegans embryogenesis, we identified a highly conserved but uncharacterized essential protein, GAD-1, that is necessary for setting cell cycle length of intestine progenitors, a function that is shared with many other factors involved in transcription, pre-mRNA splicing or polyadenylation, suggesting its potential role in mRNA biogenesis. Co-immunoprecipitation followed by mass spectrometry reveals that GAD-1 mainly interacts with non-snRNP type of splicing complex called NineTeen Complex (NTC). Consistent with this, RNA-seq analysis demonstrates pervasive defects in pre-mRNA splicing in gad-1 mutants. Transgenic reporter assay shows its ubiquitous and nuclear expression across developmental stages. Immunostaining of the C-terminal domain of RNA polymerase II demonstrates a GAD-1's role in transcription elongation. In agreement with this, depletion of GAD-1 and its interacting partners inhibits expression of both ubiquitous and tissue-specific genes, supporting that both GAD-1 and many of its interacting proteins are novel components of NTC or its associated spliceosome. Taken together, we identify GAD-1 and its multiple interacting partners as novel components of spliceosome in vivo through which they regulate pre-mRNA splicing and transcription elongation. Overall design: Mix-stage embryo samples were collected from strain BW1943 (gad-1(ct226) dpy-11(e224) V) at 16 °C and 25 °C by bleaching. Total RNAs were extracted using TRIzol (Invitrogen) following the manufacturer's instructions. mRNA purification and fragmentation, cDNA synthesis, end repairing, adapters ligation, and DNA fragment enrichment were performed using Illumina's TruSeq Stranded mRNA Library Preparation Kit based on the Kit's manual. Each library was sequenced to obtain pair-end (2×150 bps) reads using Illumina HiSeq 2500. We obtained approximately 20 million reads of high-quality score (>30 mean quality score) on average per library.