According to the United States Patent and Trademark Office, more than 3,420,560 US patents were granted between 2015 and 2025, the window this analysis covers. Research question: What share of US patents granted between 2015 and 2025 is held by the largest corporate assignees, and how steep is the concentration curve? Every figure below is computed directly from the public USPTO PatentsView dataset by the PlainPatent editorial team; see our methodology for how each number is derived and verified. Last reviewed June 2026.
The ownership curve is steep, but not as steep as you might expect
According to the United States Patent and Trademark Office, more than 3,420,560 US patents were granted between 2015 and 2025; our methodology documents how every figure here is derived (reviewed June 2026). Patent activity is often described as winner-take-all. The data tells a more layered story. Across the 3,420,560 US utility patents in our 2015–2025 window, the ten largest assignees together hold 414,936 grants — 12.1% of the total. That is a meaningful concentration for ten organizations out of more than 50,000, but it is far from monopoly. The single largest holder, SAMSUNG ELECTRONICS CO., LTD., accounts for 3.1% on its own.
Widen the lens and the curve flattens. The top 100 assignees hold 33.6% of all tracked patents, and the top 1,000 hold 61.9%. In other words, roughly a third of US corporate patenting is produced by 100 organizations, and nearly two thirds by 1,000. The remaining share is spread across tens of thousands of smaller filers — universities, mid-size manufacturers, startups, independent inventors, and government laboratories. The patent system rewards scale, yet the bulk of its participants are not household names.
Why scale compounds in patenting
Three forces push patent output toward large incumbents. The first is capital: filing, prosecuting, and maintaining a single US patent costs thousands of dollars over its life, and large portfolios require dedicated in-house counsel. The second is research throughput. Companies that already spend billions on R&D generate more patentable inventions per year, and they file defensively to protect product lines. The third is strategic stacking. In fields like semiconductors and wireless communication, firms build dense thickets of overlapping patents so that competitors cannot design around any single grant. Each of these advantages compounds, which is why the same electronics and semiconductor names recur at the top of nearly every technology class.
The flip side is that concentration varies enormously by field. A capital-intensive class such as semiconductor fabrication is dominated by a handful of foundries and chipmakers, while a broad software class can have hundreds of credible filers. Concentration is therefore a useful proxy for how contestable a technology area is. When you browse an individual technology class on PlainPatent, the leaderboard tells you immediately whether the field is locked up by incumbents or genuinely open to new entrants.
What concentration does not measure
A high patent count is not the same as high-quality innovation. Volume rewards organizations that file aggressively, but it says nothing about how often those patents are litigated, licensed, or cited by later inventions. A company with a smaller, carefully drafted portfolio in a critical technology can wield more practical power than a rival with ten times the raw count. Our Innovation Score attempts to balance volume against filing velocity, technology breadth, and claim depth precisely because raw totals overstate the lead of the largest filers.
There are also data caveats worth stating plainly. Assignee names are normalized using PatentsView disambiguation tables, which occasionally split or merge corporate families imperfectly; a parent company and its subsidiaries may appear as separate rows. Patents are counted by grant year, so the most recent year is undercounted because grants lag filing by two to three years. Provisional, design, and plant patents are excluded. None of these caveats change the headline shape of the curve, but they do mean small differences near the bottom of the ranking should be read with caution.
How to use this for your own analysis
If you are evaluating a market for competitive entry, start with the concentration of the relevant technology classes rather than the overall figure. If you are benchmarking a single company, compare its rank both overall and within its primary class — a firm can be a minor overall filer yet dominate a narrow, valuable niche. And if you are tracking innovation policy, the long tail matters as much as the top: the health of the patent system depends on whether smaller filers can still secure and defend grants against the resourced incumbents that hold the largest shares. Explore the full ranking on our top 50 companies page, and drill into any holder to see how its portfolio is distributed across technology areas.
The bottom line
Concentration in the US patent system is real but moderate. A small group of resourced corporations produces a disproportionate share of grants, yet most of the system's participants are smaller organizations operating in the long tail. The 12.1% held by the top ten is large enough to matter for any analysis of a specific technology, and small enough that the system is not closed to newcomers. The figure that best captures the structure is the top-1,000 share of 61.9%: roughly two thirds of corporate patenting comes from a thousand organizations, and the remaining third is spread across everyone else.
For practical research, treat the overall concentration number as context and the per-class concentration as the operative metric. A field where three companies hold most of the patents demands a different entry strategy than one where activity is spread across hundreds of filers, and the aggregate cannot tell those two situations apart. The value of a public dataset like this one is that it lets you move quickly from the high-level shape of the system down to the specific competitive arena you care about, with the same underlying counts the whole way down. Start broad to understand the landscape, then narrow to the class and the companies that actually bear on your question.
Methodology
We summed granted-utility-patent counts per disambiguated assignee from the USPTO PatentsView dataset (grant years 2015–2025), ranked all 50,000 assignees, and computed the cumulative share held by the top 10, top 100, and top 1,000. Counts are pre-aggregated by the PlainPatent ETL and render live from the companies table. Read the full site methodology →
Data provenance and standards
Every figure above is computed deterministically from a single ingested snapshot of the public USPTO PatentsView release, never estimated, scraped from secondary summaries, or adjusted by hand. Our pipeline disambiguates assignees, normalizes corporate naming variants, aggregates grants by year and classification, and stores the reconciled results so that any number is reproducible and auditable against the original government records. Where a measurement carries unavoidable caveats — examination lag in recent grant years, the overlapping nature of classification codes, imperfect entity reconciliation across subsidiaries — we surface those limitations explicitly rather than presenting a tidier picture than the evidence supports. This commitment to transparency, verifiable provenance, and honest uncertainty is what separates rigorous analysis from decorative statistics, and it governs every study we publish.
Primary data sources: USPTO PatentsView — granted patents; USPTO PatentsView — assignee disambiguation. Reviewed by PlainPatent Editorial · 2026-06-02.