Dinoflagellates are protists of great economical and ecological importance. Genome and transcriptome data from dinoflagellates are increasingly available, revealing extensive genomic divergence even among closely related taxa, and lineage-specific innovation of gene functions. However, most studies thus far focused on protein-coding regions in the genomes; non-coding RNAs (ncRNAs), known to play key roles in regulating gene expression in eukaryotes, remain little known in dinoflagellates. Our recent analysis of whole-genome data from dinoflagellates further revealed that non-coding regions exhibit a phylogenetic signal distinct from the coding regions. Here, we developed an analytic workflow to identify, classify, and compare polyadenylated ncRNAs using available RNA-Seq and genome data from 22 dinoflagellate taxa, including those from six genera of Family Symbiodiniaceae (Order Suessiales), Polarella glacialis (Order Suessiales), and Prorocentrum cordatum (Order Prorocentrales). We found up to 50% of the RNA-Seq reads in each transcriptome dataset mapped to putative ncRNAs in the corresponding genome. Using these data, we assessed the diversity of ncRNAs among the distinct lineages and identified the associated gene targets. Where applicable, we assessed the differential expression of ncRNAs relative to distinct growth conditions (e.g. heat stress or growth phase); for Family Symbiodiniaceae that comprise largely symbiotic species, we also assessed differential expression relative to free-living (in culture) and symbiotic (in hospite) stages. Our results reveal how ncRNAs contribute to regulation of gene expression in dinoflagellates, and provide novel insights into the molecular regulatory mechanisms that underpin diversification of these taxa.