Abstract Background Single cell sequencing of human heart tissue is technically challenging and methods to cryopreserve heart tissue for obtaining single cell information have not been standardized. Studies published to date have used varying methods to preserve and process human heart tissue, and have generated interesting datasets, but development of a biobanking standard has not yet been achieved. Heart transcription patterns are known to be regionally diverse, and there are few single cell datasets for normal human heart tissue. Methods Using pig tissue, we developed a rigorous and reproducible method for tissue mincing and cryopreservation that allowed recovery of high quality single nuclei RNA. We subsequently tested this protocol on normal human heart tissue obtained from organ donors and were able to recover high quality nuclei for generation of single nuclei RNA-seq datasets, using a commercially available platform from 10× Genomics. We analyzed these datasets using standard software packages such as CellRanger and Seurat. Results Human heart tissue preserved with our method consistently yielded nuclear RNA with RNA Integrity Numbers of greater than 8.5. We demonstrate the utility of this method for single nuclei RNA-sequencing of the normal human interventricular septum and delineating its cellular diversity. The human IVS showed unexpected diversity with detection of 23 distinct cell clusters that were subsequently categorized into different cell types. Cardiomyocytes and fibroblasts were the most commonly identified cell types and could be further subdivided into 5 different cardiomyocyte subtypes and 6 different fibroblast subtypes that differed by gene expression patterns. Ingenuity Pathway analysis of these gene expression patterns suggested functional diversity in these cell subtypes. Conclusions Here we report a simple technical method for cryopreservation and subsequent nuclear isolation of human interventricular septum tissue that can be done with common laboratory equipment. We show how this method can be used to generate single nuclei transcriptomic datasets that rival those already published by larger groups in terms of cell diversity and complexity and suggest that this simple method can provide guidance for biobanking of human myocardial tissue for complex genomic analysis.