The human genome contains about 100,000 protein-coding genes.

Before the Human Genome Project, scientists could not simply read through the complete human DNA sequence and count genes. They had to estimate the total indirectly from known proteins, RNA molecules, mutations, genome size, and comparisons with other organisms.

By the late twentieth century, estimates around 100,000 protein-coding genes were common.

A protein-coding gene is a DNA region whose information can ultimately be used to make a protein. Even with a complete genome sequence, identifying one is not always trivial: researchers must determine where genes begin and end, which RNA transcripts are real, and whether a predicted sequence actually produces a functional protein.

When the first draft of the human genome appeared in 2001, the count was already dramatically below 100,000. Later annotation brought the estimate down toward roughly 20,000 protein-coding genes.

The surprise showed that biological complexity does not come simply from having vastly more protein-coding genes. Cells can use the same genes in different ways through regulation, alternative RNA splicing, chemical modification of proteins, and networks of interacting genes.

The exact number continues to be refined as small proteins and gene boundaries are reassessed.

This count also refers specifically to protein-coding genes. The human genome contains many additional genes whose functional products are RNA rather than proteins.