DRAFT BUILD · hub-companies · 0 of 7 blocks still have no text · noindex
AmericaPorts.com
HomeCompanies

Directory · two metro counts across eleven ports · register read 07/28/2026

Licensed intermediaries are counted here by the place they give an address in, and the national total everyone quotes does not exist

⚑ The short answer
  1. Licensed intermediaries are counted on this site by the place they give an address in, and are never ranked, recommended or listed by name. That is a decision about what this catalogue covers: ports, terminals and gateways, with a company appearing only as an attribute of a place. No company is named anywhere on this page.
  2. Ocean transportation intermediaries number 9,795 organisations in FMC's own register download, and 5,205 of them hold a licence number. The two lists the register publishes on its site cannot be added, because 2,588 organisations carry both roles — an overlap the file measures and the web counters cannot. FMC publishes no aggregate of its own; the count above is ours, taken from their file.
  3. US container ports are the unit of this catalogue and a place name is the unit of the register, and the two do not line up. The 95 organisations counted across Los Angeles, Long Beach, San Pedro, Wilmington and Terminal Island cover two ports here; the 11 at Savannah, Pooler and Garden City cover one; and eight of the eleven ports profiled on this site have not been counted at all.
  4. Registers cover one leg of a route each, and we hold only one of them. The FMC list covers the ocean leg, the road leg belongs to a different regulator and we have captured nothing from it, and at one port the party billing for the days a box sits is the port authority itself, which no intermediary register covers in either direction.
No company is named anywhere on this pageTwo federal lists, and they cannot be addedEight of eleven ports carry no count at all

How many licensed organisations give an address at each port

Licensed intermediaries are counted on this site by the place names a port’s own berths are addressed at — five of them at San Pedro Bay, three at Savannah — because the register records an address and the only honest boundary is one we can print. The two counts do not line up with our own unit either: one figure covers two ports at once, and eight of the eleven ports profiled here carry no count at all.

How this index is built · what we are paid

  • No company is named, listed, linked or characterised anywhere on this page. This site covers places — ports, terminals and gateways — and an organisation appears here only as a count of records at a place. Nobody can pay to appear in this index, because nobody appears in it.
  • Nothing here is sorted, and the reason is stronger than on the other hubs. Ordering ports by how many organisations give an address near them would read as a ranking of the ports, and the number cannot support one.
  • The register records an address, not a route. An organisation counted at a port has not been shown to move a single box through it, and one that moves cargo through it every week may be addressed in another state.
  • Every count on this page now comes out of one file — the register FMC publishes for download, stamped 28 July 2026 and counted end to end on 31 July. The web counters read two days earlier are gone from the page; one of them was not even the same unit.
  • Both counts below print the boundary they were drawn with, as a list of place names, and both come back out of the file when you apply it. A count of a place whose edge is not written down is not a figure anyone can check, including us — which is how the previous figure for the California rows was lost.
  • No port, terminal, carrier or intermediary has paid to appear here or to be left out, and no link on this site is a referral link.
How many licensed organisations give an address at each port this site covers.
PortOrganisations addressed thereWhat the register actually countedWhat the number does not tell you
Los Angeles95Los Angeles · Long Beach · San Pedro · Wilmington · Terminal Island, in California — the boundary, printed, and the figure comes back out of the fileorganisations addressed at the five place names the two ports’ own berths use — one figure covering both harbours and so both rows of this cataloguenothing about any single organisation, and nothing about which port any of them works through
Long Beach95the same figure as the row above: the basket covers both harbours, so it belongs to both rows or to neither, and it is counted once · Los Angeles · Long Beach · San Pedro · Wilmington · Terminal Island, in California — the boundary, printed, and the figure comes back out of the filethe same five place names, which the register does not divide between the two harboursnothing that would let you split the figure between the two ports, and nothing about any single organisation
Savannah11Savannah · Pooler · Garden City, in Georgia — the boundary, printed, and the figure comes back out of the fileorganisations addressed at the three place names that port’s facilities use, rather than at a portthe cleanest match between the register’s boundary and ours, and even here the boundary is not a port
Houstonnot checked yetno basket of place names has been drawn for this port, so nothing has been counted from the fileno count is estimated from a neighbouring port, and none is inferred from the ones above
The other seven port profilesnot checked yetnone of the seven has been counted; for one of them — a port made of two harbours — counting would first mean deciding whether the two are one market, which this site leaves opentheir absence from this column is our search, not an absence of organisations addressed there
Two rows share one figure and it is counted once. The 95 covers a basket containing both harbours, so it belongs to both rows or to neither — printing it in each is a statement about the boundary and not two separate findings. Neither number here is a metro count. They are counts of named places, and the names are printed beside them precisely so that nobody has to take the edge on trust. The national totals are not in this table and cannot be built from it; they are the subject of the block below.

FMC OTI List · file stamped 2026-07-28 13:30:02 EST · counted 07/31/2026 by build_fmc.py

What an empty-looking cell means: not checked yet — we have not read the document yet
What this column measures, and what it cannot. Roughly how much intermediary activity is addressed at those place names — that is the whole of it. It is not a measure of choice, of quality, of capacity or of price, and two ports with different counts have not been shown to differ in anything a shipper can act on. A count this small at one port and this large at another says where organisations keep their letterhead, not where cargo moves.

Every US container port this catalogue covers, including the eight carrying no count: every US container port we cover.

What to do with a count, and what to do instead

There are two things a reader arrives here wanting, and this catalogue can do one of them.

From a port, the count above tells you roughly how much intermediary activity is addressed at those place names. That is the whole of what it tells you, and the caveats around it are not fine print — they are the meaning of the number.

From a company name, this catalogue cannot help and will not pretend to. No organisation is named anywhere on this site. The federal register is public and searchable by name, and that is where a specific licence gets checked — by whoever is about to hand that company their cargo.

The question worth asking is a different one: not whether a licence exists, but which leg of your route it covers. A genuine licence can cover a leg the company does not run, and that is a difference in what is regulated rather than a sign that anything is wrong.

Half your route has no register in our reading at all. A licence from this register covers the ocean leg. The road leg — the largest unpriced piece of an inland routing — is licensed by a different federal agency, and nothing from its records has been captured here: that is the single gap declared on this page. At one Gulf port the party billing for the days a box sits is the port authority itself, which no intermediary register covers in either direction. A reader asking who is licensed to handle their box gets an answer about one leg of two, and this is the site saying which one.

What a licence covers, and the parts of your bill it does not

An entry in this catalogue holds three things — a count, the list of place names it covers and the date it was taken — and nothing else. What the licence behind those records covers is the more useful question, and it has a narrow answer.

The federal rules you can actually act on are not about the licence. They are about the invoice: what a demurrage bill has to contain and who may be billed for it. That rule was codified in 2024 and has been before the D.C. Circuit twice since, which is why this site publishes the codification and the fact of the litigation rather than a recitation of its conditions: the federal rules your invoice has to meet.

Why there is no national total, and what we counted instead

Licensed intermediaries have no national total that FMC publishes. The register keeps two lists, prints a result count on each and publishes no aggregate over them — which is how a figure close to nine thousand came to circulate as though it were one.

Their own download settles it. The file FMC publishes carries one row per organisation, and all 9,795 organisation numbers in it are distinct. The roles sit in separate columns, so the overlap that made the two lists unaddable stops being unknown and becomes a measurement: 2,588 organisations hold both. Counting licence numbers directly, without summing anything, 5,205 organisations hold one.

That is a count of ours and not of theirs, and it is a count of organisations rather than of licences — 5,205 organisations carry only 5,191 distinct numbers between them, because fourteen numbers each sit on two records.

The FMC register recounted from its own download, one row per organisation.
WhatHow manyWhat that means
Organisations in the register9,795the whole file, stamped 28 July 2026 and counted 31 July
Holding a licence number5,205675 forwarder-only, 1,942 NVOCC-only and 2,588 carrying both roles — the count FMC does not publish and its own file yields
Registered without a licence4,5904,588 of them addressed outside the United States; this is what makes the NVOCC list mixed
Carrying the freight forwarder role3,263every one licensed. The web counter said 3,273 two days earlier — and counted trade names rather than organisations, so it was never the same quantity
Carrying the NVOCC role9,120of which 4,530 licensed. The web counter said 9,030 two days earlier, in the same unit
Distinct licence numbers behind the 5,2055,191fourteen numbers each appear on two organisation records, so licences and licensees are not one count either
The figure everyone quotes is not a total, and we can now say what it is instead. A round number close to nine thousand circulates as the count of licensed US ocean intermediaries. It is the size of the NVOCC list — 9,120 organisations in the download — and fewer than half of those hold a licence: 4,530 do. As a count of licensed intermediaries it is roughly double. FMC still publishes no aggregate of its own, so the 5,205 above is labelled as ours throughout, and it is reproducible from the same file by anyone who downloads it.
Both port counts now print the boundary they were drawn with, and one of them had to be drawn again. Savannah always did: filter the file to Savannah, Pooler and Garden City and eleven come back. The California figure did not — it was published as a metro count with no metro written down anywhere, and no boundary reproduces it: whole-county gives 1,044, the whole state 1,328, the two port cities alone 86. So it was not reverse-engineered to fit. The same rule was applied as at Savannah — the place names the port’s own berths are addressed at — and that basket yields 95, not 464. The old figure is in the corrections log with the number it replaced.
The largest concentration in California is not at the water. One inland city forty kilometres from the berths carries 92 records in this register, against 95 for both harbours and their three port place names put together. The register records where an organisation keeps its letterhead, not where cargo moves, and this is the sharpest single illustration of it in the file. It is also why a bigger boundary would not have made the port figure more meaningful — only larger.

What each port is willing to publish about itself, measured the same way this page measures itself: what each port is willing to publish.

How this page is built, and why it names no company

How every figure here is sourced, what each status means, and the corrections log: how every figure here is sourced.

Questions about licensed intermediaries and US ports

How many licensed ocean intermediaries are there in the United States?

FMC publishes no such total, but its own downloadable register lets one be counted, and on the file of 28 July 2026 it is 5,205 organisations holding a licence number out of 9,795 in the register. That count is ours rather than theirs, and it counts organisations rather than licences — only 5,191 of the numbers are distinct.

Is the figure of about nine thousand licensed intermediaries correct?

No. It is the size of the NVOCC list, which the download puts at 9,120 organisations, and only 4,530 of those hold a licence — the rest are registered foreign NVOCCs without one, which the register states on the same page. As a count of licensed intermediaries it is roughly double.

Which companies are licensed at my port?

This site does not name companies. It covers ports, terminals and gateways, and an organisation appears here only as a count at a place. The register itself is public and searchable, and that is where a specific name is checked.

What does the count at a port actually tell me?

Roughly how much intermediary activity is addressed at those place names, and nothing more. The register records an address rather than a route: one inland California city carries 92 records against 95 for both harbours together, which is the whole caveat in one comparison.

Why do Los Angeles and Long Beach share one number?

Because the boundary is a set of place names covering both harbours — Los Angeles, Long Beach, San Pedro, Wilmington and Terminal Island — and every one of the 95 organisations sits inside it. The figure belongs to both rows or to neither, so it is shown once with both ports named.

Why is there no count for Seattle and Tacoma?

Counting a port made of two harbours would mean first deciding whether the two are one market, and that is a question this site leaves open. Rather than settle it inside a count, no count is attempted there.

Does a licence mean the company can handle my whole shipment?

No. An OTI licence covers ocean transportation intermediation. The road leg is licensed by a different regulator whose records we hold nothing from, and a terminal's storage charge and a port authority's own fee sit outside any intermediary register. A genuine licence can cover a leg the company does not run.

A company is not in the register — is that a warning sign?

Not by itself. Absence is not a finding: an organisation may be addressed elsewhere, hold a different kind of authority, or be a party the register does not cover at all. The question worth asking is which leg is covered by whose licence.

How current are these numbers?

The register counts come from FMC's own download stamped 28 July 2026 and were counted end to end on 31 July. They replace the web counters read on 26 July, which answered a slightly different question: the forwarder page counts trade names and the file counts organisations. The per-port counts come from the same file.

Can I reproduce these numbers myself?

Yes, all of them. The register counts come from the file FMC publishes for download, one row per organisation number with the role and licence columns read as they stand. Both port counts print the list of place names they were drawn with, so filtering the same file the same way gives the same 95 and the same 11.