What is the difference between PSTN and PBX?

The clearest way to separate the two is by who each one serves. The PSTN, short for public switched telephone network, is the worldwide collection of interconnected telephone networks run by carriers, and its job is to connect any phone to any other phone across providers and countries.[4][5] A PBX, short for private branch exchange, is a switching system owned by a single organization that connects that organization's own phones to each other and gives them shared access to the outside world.[1][2] The PSTN is public infrastructure that everyone shares, while a PBX is private equipment that one business runs for its own staff, so they sit at very different scales even though they plug into each other.

The name of a PBX spells out the relationship. 'Private' means the system stays separate from the public network even though it connects to it, 'branch' refers to the way a smaller endpoint hangs off the main circuits, and 'exchange' describes the switching that lets many phones share a limited set of lines.[1] Before PBXs were common, a company that wanted ten desk phones to reach outside callers would have needed ten separate lines running to the telephone company's central office. A PBX removes that one-to-one requirement by concentrating traffic, so a business with 100 employees might run on roughly ten external lines, because most calls stay inside the building and the outside lines get used only when someone actually dials out.[2][3]

The two therefore solve different problems. The PSTN handles the routing that gets a call from one network to a phone on another, across regions and carriers, and a PBX handles the local routing inside one site plus the sharing of a handful of outside lines.[1][5] They meet at the edge of the premises, where trunk lines run from the PBX out to the PSTN, and beyond that boundary the public network takes over.[1][4] The sections that follow break the differences down one at a time, then explain how the two connect and where cloud systems are blurring the old line between them.

Ownership is the root difference, and it drives almost everything else. The PSTN is public infrastructure, built and run by national, regional, and local telephone operators that interconnect their networks so a call can cross from one carrier to another and reach a phone almost anywhere.[5] No single company owns it, and you connect to it as a subscriber by buying service from a carrier rather than by installing your own switch.[4] A PBX sits at the other end of the ownership scale: one organization owns or subscribes to it and runs it for its own people, which is what the word 'private' in private branch exchange means.[1] Historically that meant a switch in a back room that the company or its vendor maintained, and today it can also be software rented from a hosted provider.[2] The consequence is control. Because a business runs its own PBX, it decides the dial plan and how calls flow internally, along with which calling features to turn on, whereas the behavior of the public network is set by carriers and regulators.[1]

The two switch calls at very different scopes. A PBX switches calls between the phones at one location and connects those phones to a limited number of external lines, so an internal call from one desk to another never leaves the building or touches a carrier.[1] That local switching is where a PBX earns its keep, because internal calls cost nothing and only outbound calls use up a shared outside line.[3] The PSTN switches at a far larger scope, routing calls between exchanges, between carriers, and between countries through a hierarchy of switches so that any subscriber can reach any other.[5] A useful way to picture it is that a PBX is a small, private version of what a public telephone exchange does, scoped to one organization rather than a whole town or region.[1][5] When someone at a PBX phone dials an outside number, the PBX seizes an available trunk and hands the call to the PSTN, which carries out the wider routing to the destination.[1]

Location is the difference most people can point to. A traditional PBX is physical equipment on the customer's premises, historically a cabinet where the incoming trunk lines terminate and connect to the internal phones.[1] The organization houses it, powers it, and maintains it, which is why an on-premises PBX is a capital purchase as much as a service.[2] The PSTN lives in the carrier's world instead: telephone exchanges, the trunks between them, and the long-haul network that spans regions, none of which sits inside the customer's building.[5] Hosted and cloud PBX systems have shifted this picture, because the switching software runs in a provider's data center and the business keeps only the phones or apps, paying a per-user subscription rather than buying hardware.[2] The switching logic stays the same; only its physical home changes, which is one reason the boundary between owning a PBX and buying a service from the network has softened.

The PSTN and a PBX identify phones with different kinds of numbers. Every phone reachable on the PSTN has a full public number under the international numbering plan defined in ITU-T Recommendation E.164, which gives each line a globally unique, routable address of up to 15 digits.[8] Inside a PBX, phones are usually reached by short internal extensions, often three or four digits, that only carry meaning within that system.[1] The two schemes are bridged by a carrier feature called direct inward dialing, or DID. With DID, the telephone company allocates a block of public numbers to the organization and delivers them over a smaller set of trunks, and the PBX maps each incoming public number straight to the right extension without an operator.[6] This is how a company can publish a distinct outside number for every employee while running on far fewer physical lines than it has numbers.[6] Internally, staff still dial the short extension; externally, callers reach the same person through the full E.164 number that DID points at that extension.[6][8]

The cost models differ because one is a service and the other is a system you run. PSTN access is billed by a carrier, traditionally per line and per call or per minute, so you are paying for connectivity to the public network as an ongoing service.[4] A PBX changes what you are buying. Instead of a line per person, you buy or subscribe to a switch and a shared set of trunks, then pay the carrier only for those trunks and the calls that use them.[1] An on-premises PBX has historically been a capital cost, the equipment plus its maintenance, while a hosted or cloud PBX turns that into a predictable per-user monthly fee with the provider absorbing the hardware.[2] The saving that made PBXs standard is straightforward: sharing a small pool of outside lines across many phones costs far less than giving every phone its own line to the central office.[1][3]

A PBX reaches the PSTN through trunk lines, the shared external circuits that carry calls between the private switch and the public network.[1] When an internal caller dials an outside number, the PBX selects a free trunk, passes the call to the PSTN, and the public network handles the routing to the destination.[1][3] For decades those trunks were physical carrier circuits: analog central office lines for small systems, or a digital Primary Rate Interface delivering multiple channels over a T1 or E1 for larger ones.[6] The number of trunks sets how many outside calls can run at once, which is why capacity planning for a PBX is about concurrent calls rather than headcount.[1]

Most new deployments now use a SIP trunk instead of a physical circuit. A SIP trunk carries calls as IP data to a provider that connects them to the PSTN, replacing the older analog line or PRI while doing the same job of linking the PBX to the public network.[2][10] Incoming calls travel the same path in reverse. The PSTN delivers a call to the organization's trunks, and direct inward dialing hands it to the correct extension, so an outside caller reaches an individual desk without going through a switchboard operator.[6]

The clean split between a private box and the public network has softened, because the PBX has moved into software and the cloud. An IP PBX performs the same call switching as a traditional PBX but uses digital IP signaling over the office data network rather than analog lines, so phones connect with Ethernet instead of dedicated telephone wiring.[1][2] A hosted or cloud PBX goes further, moving the switching itself into a provider's data center; the business keeps only handsets or software apps and pays a per-user subscription, while the provider runs the equipment and its connection to the PSTN.[1][2] The underlying distinction still holds. A private switching layer, wherever it physically runs, serves one organization and hands its outside calls to the shared public network, which is the same relationship a copper-era PBX had.[2][5]

Choosing how to handle business calling is really a question of where you want the switching to live and how you want to reach the PSTN, and each option carries a different set of practical concerns.

On-premises PBX

Running your own PBX gives the most direct control over the dial plan and features, and internal calls stay entirely on your equipment.[1] The cost is that you own the hardware, its maintenance, and the trunks, and you carry the risk if the switch or its power fails.[2] It suits organizations that want that control and have the staff to manage it.

Hosted or cloud PBX

Moving the PBX to a provider removes the hardware and shifts to a predictable per-user fee, with the provider handling upgrades and the PSTN connection.[2] The trade-off is that call quality and uptime now ride on your internet connection and the provider rather than on a dedicated telephone circuit.[4] It fits organizations that prefer an operating cost and remote-friendly access over on-site control.

The PSTN transition

The public network itself is changing underneath all of these choices. Carriers are retiring the legacy circuit-switched PSTN in favor of IP, and regulators describe the old analog network as having reached the end of its serviceable life, with landline service moving onto broadband-based voice.[4][9] The practical effect is that whatever PBX you run, its link to the outside world is increasingly a SIP or IP trunk rather than copper, so planning around IP is now the safer default.[9][10]

A few common mix-ups blur these two, and separating them clarifies both.

A PBX is not the PSTN

A PBX is a private switch for one organization, not the public network, even though it connects to it.[1][5] It routes your internal calls and shares your outside lines, while the PSTN carries calls between everyone, across carriers and borders.[5]

Neither one is VoIP by itself

Voice over IP is a method of carrying calls as data, not a network or a switch.[4] A PBX can be analog or IP, and the PSTN is being rebuilt on IP, so 'VoIP' describes how calls travel rather than naming either the private switch or the public network.[2][4]

A SIP trunk is not a PBX

A SIP trunk is the connection that links a PBX to the PSTN over IP, not the switching system itself.[1][10] The PBX does the routing; the SIP trunk is one way it reaches the outside world, in the same role a PRI or analog line once filled.[6][10]

Centrex is not an on-premises PBX

Centrex offers PBX-like features but runs from the carrier's central exchange on the carrier's own equipment, so a business using it has no switch of its own on site.[7]