Single-cell gene expression assays are critical tools for identifying functional subtypes of immune cells, as well as changes within cells resulting from pathology or treatment. However, most current methods are expensive, low in sample throughput, rely on reverse transcription and lack the sensitivity to consistently measure low expressed genes from single cells - preventing measurements of many key immune biomarkers and molecular pathways. Another problem with existing single-cell sequencing methods is that many of the cell-associated reads either fail to map to transcripts or map to uninformative ribosomal and mitochondrial genes. Some methods also require the use of live cells, which means the single-cell assay has to be carried out at the specific time of treatment or when culture is terminated. The aforementioned limitations have contributed to most single-cell sequencing studies being restricted to small numbers of samples and have prevented their wider scale adoption in biopharma, clinical, translational and applied applications. Here we report on the development and performance of a novel combinatorial split-pool-based workflow, TempO-LINC™, that adds cell-identifying molecular barcodes onto high-sensitivity gene expression probes within fixed cells. Cells can be fixed at the end of a treatment, or when obtained from a patient and stably stored for months until they are ready to be processed with other samples through the TempO-LINC assay. All probes within the same fixed cell receive an identical barcode, enabling the reconstruction of single-cell gene expression profiles across as few as several hundred cells and up to 100,000 cells per run. TempO-LINC simultaneously processes up to 96 fixed cell samples per run, has a simple workflow that is amenable to automation and as we will show, has a higher gene detection rate than existing single-cell platforms. We will also show that TempO-LINC has an observed doublet rate of less than 0.5% for 12,500 cells analyzed. Critically, TempO-LINC reduces sample sequencing costs and although we show the assay accurately profiles the whole transcriptome (23,000 transcripts), it can also be targeted to measure only genes/pathways of interest - dramatically reducing sequencing costs yet further. We applied TempO-LINC to assay PBMCs and a human hepatocyte model for toxicology and demonstrated the ability to correctly identify unique cell states and molecular pathways from these populations. We will also present time course and dose response data from compound treatment of cells. All of these features position TempO-LINC as a potential disruptive technology that hematology researchers, clinicians and drug developers can use for new large-scale applications/studies using single-cell sequencing across thousands of samples while obtaining improved data quality.