What is the average size of a protein?

The practical answer is about 300 amino acids for a bacterial protein and about 400 amino acids for a eukaryotic protein. In mass terms, that is roughly 30–50 kilodaltons. A human ribosome can build such a protein in about one minute.

Those numbers describe a typical protein, not every protein. The range is enormous, from about 20 amino acids at the small end to more than 45,000 at the large end, and it differs systematically between the domains of life.

What is the typical size of a protein?

The median protein in bacteria is about 267 amino acids long. In archaea, it is about 247 amino acids. In eukaryotes, the median rises to 361 amino acids.^[1]

Converted to mass, those medians are approximately 30 kilodaltons (kDa) for bacteria and archaea, and about 40 kDa for eukaryotes. The conversion uses the average molecular weight of one amino acid, roughly 110 daltons, after accounting for the water molecule lost in each peptide bond.^[5]

The full distribution is broad and right-skewed. Most proteins cluster between 100 and 500 amino acids, but long tails extend to many thousands of residues in all domains of life. Because of this skew, the arithmetic mean is usually larger than the median - often 10–20% higher - which is why different sources report slightly different "average" values for the same proteome.^[1][2]

In the human proteome, the median protein length is about 375 amino acids, corresponding to about 42 kDa. If you randomly select a protein molecule from a human cell rather than a gene from the genome, the abundance-weighted average is similar, roughly 400 amino acids, because highly expressed proteins like ribosomal proteins and metabolic enzymes fall near the middle of the size distribution.^[1][2]

How does protein size differ between bacteria and eukaryotes?

Eukaryotic proteins are consistently longer than bacterial and archaeal proteins. The difference is not subtle - the median eukaryotic protein is roughly 35% longer than the median bacterial protein.^[1]

This holds across essentially all functional categories. Enzymes, structural proteins, signaling proteins, and transcription factors are all longer in eukaryotes than their counterparts in prokaryotes. The difference is not explained by a few outlier protein families - it appears in the vast majority of orthologous groups.^[1]

The main explanation is domain accretion. Over evolutionary time, eukaryotic proteins acquired additional domains - modular folding units that add new functions - through gene fusion and exon shuffling. A eukaryotic protein often combines what would be several separate proteins in a bacterium into a single, longer polypeptide chain.^[1]

Eukaryotic proteins also tend to have longer intrinsically disordered regions - stretches that do not fold into a fixed three-dimensional structure but contribute to signaling, regulation, and protein-protein interactions. These regions add length without adding ordered domains.

Not all eukaryotes have long proteins. The yeast Saccharomyces cerevisiae has a median of about 379 amino acids, while the plant Arabidopsis thaliana has a median of about 356. Among bacteria, endosymbionts and parasites with reduced genomes tend to have shorter proteins, while free-living species like Escherichia coli have a median of about 278 amino acids.^[1]

What are the smallest and largest proteins?

At the lower end, the smallest functional proteins are about 20–30 amino acids long. Naturally occurring zinc-finger domains can be as short as 25 residues and still fold into a stable structure. Designed mini-proteins such as Trp-cage are stable at just 20 amino acids.^[4]

Short ribosomal proteins of about 100 amino acids are common, and many peptide hormones and toxins are even smaller - some under 15 residues - though whether these are considered "proteins" or "peptides" is a matter of definition. The conventional boundary is roughly 30–50 amino acids.

At the upper end, the largest known protein is PKZILLA-1, a giant polyketide synthase discovered in 2024 from the alga Prymnesium parvum. It spans 45,212 amino acids with a mass of 4,761 kilodaltons, about 120 times the length and mass of a typical human protein.^[3]

PKZILLA-1 overtook the previous record-holder, titin, which is the largest human protein. The canonical human titin isoform is 34,350 amino acids long and serves as a molecular spring in muscle sarcomeres. The titin gene (TTN) encodes up to 38,138 residues in its longest isoform.^[6]

Two other notably large proteins are dystrophin (3,685 amino acids), which links the muscle cell cytoskeleton to the surrounding membrane, and nebulin (6,669 amino acids), another giant muscle protein. Even these, however, are an order of magnitude smaller than titin and PKZILLA-1.

How long does it take to make a protein?

A human ribosome adds about 10 amino acids per second to a growing protein chain. At that rate, an average 375-amino-acid human protein takes roughly 40 seconds to synthesize. The practical number cited in educational sources is about one minute for an average-sized protein, accounting for initiation and termination time.^[4]

In bacteria, translation is faster - roughly 15–20 amino acids per second - so a typical 270-amino-acid bacterial protein is made in about 15 seconds.

Multiple ribosomes can read the same messenger RNA simultaneously, spaced about every 100–200 nucleotides along the transcript. A single mRNA can produce about 140 protein copies per hour in human cells.^[4]

For the largest proteins, synthesis times are much longer. At 10 amino acids per second, building a single copy of the 45,212-residue PKZILLA-1 would take over an hour of continuous translation. A titin molecule would take roughly 55 minutes. This is part of why giant proteins tend to have specialized functions in cells that do not rapidly divide - the synthesis cost is high.

Protein length is one scale question. Protein sequence counts describe how many distinct sequences have been catalogued, while DNA length describes the genome sequence that encodes them.

Methodology

The median protein lengths reported here come from Brocchieri and Karlin's 2005 analysis of whole-genome proteomes from 5 eukaryotic, 16 archaeal, and 67 bacterial species. These numbers remain the most widely cited reference values, and subsequent analyses of larger datasets have not substantially changed the central estimates.^[1]

Mass estimates are derived from sequence length using the standard conversion of approximately 110 daltons per amino acid residue, the weighted average molecular weight of the 20 standard amino acids as they occur in natural proteins.^[5] The abundance-weighted distribution referenced in the text comes from Milo and Phillips, who integrated genome sequences with proteome-wide mass spectrometry censuses.^[2]

The PKZILLA-1 size comes from Fallon et al. (2024); the titin size from the UniProt entry for human TTN; translation speed and cell-level estimates from the 2025 NIGMS educational summary.^[3][4][6]

Citation

ProteinIQ. "What is the average size of a protein?" Updated June 30, 2026. Accessed [your access date]. https://proteiniq.io/guides/average-protein-size

Sources▼

1. Protein length in eukaryotic and prokaryotic proteomes Nucleic Acids Research · 2005. https://pmc.ncbi.nlm.nih.gov/articles/PMC1150220/
2. Cell Biology by the Numbers - How big is the average protein? Garland Science · 2015. https://book.bionumbers.org/how-big-is-the-average-protein/
3. Giant polyketide synthase from Prymnesium parvum (PKZILLA-1) Science · 2024. https://www.science.org/doi/10.1126/science.ado3290
4. Proteins by the Numbers National Institute of General Medical Sciences · 2025. https://nigms.nih.gov/biobeat/2025/01/proteins-by-the-numbers/
5. ExPASy ProtParam documentation - average amino acid molecular weight SIB Swiss Institute of Bioinformatics · June 30, 2026. https://web.expasy.org/protparam/protparam-doc.html
6. Titin - human TTN gene and protein UniProt · June 30, 2026. https://www.uniprot.org/uniprotkb/Q8WZ42/entry