A comprehensive cellular anatomy of normal human kidney is crucial to address the cellular origins of renal disease and renal cancer. Some kidney diseases may be cell type-specific, especially renal tubular cells. To investigate the classification and transcriptomic information of human kidney, we performed a method to obtain single-cell suspension of kidney rapidly, and conducted single-cell RNA sequencing (scRNA-seq). We present scRNA-seq data of 23,366 high quality cells from human kidneys of 3 donors. In this dataset, we show 10 clusters of normal human renal cells. Due to the high quality of single cell transcriptomic information, proximal tubule (PT) cells were classified into 3 subtypes and collecting ducts cells into 2 subtypes. Collectively, our data will provide a reliable reference for the studies of renal cell biology and kidney diseases. Overall design: Primary kidney samples from 3 patients were used for single cell RNA sequencing by 10X Genomics Please note that both raw and processed data files associated with each sample have been updated on Oct 31, 2019.