We present Single-cell TOtal RNA Miniaturized sequencing (STORM-seq), a unique molecular identifier (UMI)-containing, full-length single-cell ribo-reduced RNA sequencing protocol, optimized to profile thousands of cells per run. Using random hexamer priming, STORM-seq recovers comprehensive RNA profiles from single cells with library complexity approaching that of bulk RNA-seq. STORM-seq identifies thousands of additional coding and non-coding transcripts not detected by existing methods, and recovers clinically relevant structural variants and unbiased transposable element expression at the single-cell level. Incorporation of ERCC spike-ins facilitates proper normalization and cross-experiment comparisons, and preserves biologically meaningful intercellular differences in total RNA abundance. Importantly, STORM-seq is characterized by simple library preparation and bioinformatics procedures, and can be executed using standard laboratory equipment and off-the-shelf reagents, thus rendering it widely accessible. We apply STORM-seq to primary human fallopian tube epithelium (FTE), revealing intermediate/transitional cell states, and a putative progenitor cell population. The results support a trajectory from a multipotent progenitor population to ciliated and secretory cell types in normal FTE. These findings are consistent across human subjects, sequencing depths, and platforms. Long intergenic non-coding RNAs (lincRNAs) appear as key driver genes in both ciliated and secretory lineage trajectories, underscoring the importance of both coding and non-coding RNA in this tissue. Additionally, we observe coordinated transposable element expression, shaping the ciliated lineage trajectory. By capturing essentially complete individual cellular transcriptomes, STORM-seq sheds new light on the transcriptional programs that establish cellular state and fate in complex tissues.