Genome-wide association studies have identified numerous single-nucleotide polymorphisms (SNPs) associated with human diseases or phenotypes. However, causal relationships between most SNPs and the associated disease have not been established, owing to technical challenges such as unavailability of suitable cell lines. Recently, efficient editing of a single base pair in the genome was achieved using programmable site-specific nucleases. This technique enables experimental confirmation of the causality between SNPs and disease, and is potentially valuable in clinical applications. In this review, I introduce the molecular basis and describe examples of single-base pair editing in human cells. I also discuss the challenges associated with the technique, as well as possible solutions.