What do the letters at the start of a tracking number mean?

The first two letters of a 13-character postal tracking number are a service indicator, and they name the postal product rather than the carrier, the destination or the delivery speed. They are defined by Universal Postal Union standard S10, which fixes the identifier at two letters, eight digits, one check digit and a two-letter country code. R covers registered letter post, C covers parcel post, E covers EMS, L covers tracked letter post and V covers insured letter post. A number that opens with 1Z, or with digits alone, is not an S10 identifier at all and follows the issuing carrier's own numbering scheme.

Who assigns the two letters, and under which standard

Service indicators are allocated under the authority of the UPU Standards Board and issued nationally by one assigned organisation per member country. The current specification, S10 version 12, was approved on 17 October 2017 and reached UPU status 2 on 10 May 2011. It applies to the item classes that require individual tracking: EMS items, parcels, and registered, recorded delivery, tracked and insured letter-post items. The same 13 characters are carried on UPU forms such as the CN 33, CN 16 and CP 87, and on the CN 08 inquiry form, so the prefix printed on a posting receipt is the string that later drives a claim.

"For service indicators intended for universal use between designated operators, the first character indicates the type of postal product or service and the second character is assigned by the origin operator from a range of allowed values." (Universal Postal Union, standard S10-12, clause 5.2, approved 17 October 2017.)

Two letters are used rather than one because the pair carries two different kinds of information. The first letter is a global classification that every designated operator reads the same way. The second is a national allocation, which is why a reader can infer the product class from a prefix with confidence but cannot infer the exact service name.

The service indicator ranges, product class by product class

S10 clause 5.6 assigns nine ranges for universal use and eight further ranges for domestic, bilateral or multilateral use only. The ranges below are the allocation as published in version 12 of the standard; the sample identifiers are the standard's own illustrations and are fictitious rather than live numbers.

Range	Product type	Notes and illustrative example
EA to EZ	EMS	EX to EZ require bilateral agreement. Example: EE123456785KR
LA to LZ	Letter post, tracked	LZ requires bilateral agreement
MA to MZ	M bags	Bulk despatches of printed matter to one addressee
QA to QM	IBRS, International Business Reply Service	13 combinations only, the narrowest universal range
RA to RZ	Letter post, registered	RZ requires bilateral agreement. Example: RR876543216ER
UA to UZ	Letter-post items containing goods, without customer-facing tracking	Used for supply-chain visibility and advance electronic data
VA to VZ	Letter post, insured	VZ requires bilateral agreement. Example: VA456789015KG
CA to CZ	Parcel post	CZ requires bilateral agreement. Examples: CP654321092GM, CV010000155UA
HA to HZ	ECOMPRO parcels	HX and HY require multilateral agreement, HZ bilateral
AV to AZ, BA to BZ, DA to DZ, GA, GD, NA to NZ, PA to PZ, ZA to ZZ	Domestic, bilateral or multilateral use only	AV to AZ is intended in principle for RFID-tracked items

Where an item could fall in more than one range, the standard sets an order of precedence for letter post: insured wins, then registered, then tracked, then M bag, then IBRS, and an item that fits none of those but contains goods takes a U range indicator. That precedence explains a result that surprises senders. An insured registered letter carries a V prefix and not an R prefix, because insurance outranks registration. The difference between registered and ordinary mail is therefore visible in the first character before any scan event exists.

Why the same service starts with RA in one country and RR in another

Only the first character is standardised, which leaves 26 possible second characters inside every product range. An operator picks its second letters to separate its own products, contract customers or serial blocks, and those choices are not published in a single global list. RR and RA are both registered letter post, and the difference between them is a national decision, not a difference in service level that a foreign reader can decode.

This is where most prefix guides go wrong. A statement such as "RR means registered post and RA means a small registered packet" can be accurate for one operator and wrong for the next, because both belong to the same RA to RZ range and the standard assigns no meaning to the second letter. The dependable reading is the first character alone: R means the item is registered, whatever the second letter says. The same caution applies to CP against CV, where the standard notes that insured parcels may carry an indicator other than CV, and that CV does not by itself imply an insured parcel.

For items with no prefix at all the question is different, because untracked letter post carries no S10 identifier: ordinary airmail is not trackable for that reason rather than through any failure of the network.

The eleventh character, and how to test a number for typing errors

Character 11 of an S10 identifier is a check digit computed from the eight serial digits with a weighted modulus 11 algorithm. The weights, applied left to right, are 8, 6, 4, 2, 3, 5, 9 and 7. The weighted values are summed, the sum is divided by 11, and the remainder is subtracted from 11: a result of 1 to 9 becomes the check digit, a result of 10 becomes 0, and a result of 11 becomes 5.

"A check digit is an arithmetically derived number that is used to support the detection of substitution and transposition errors which can occur during data capture." (Universal Postal Union, standard S10-12, clause 5.4, approved 17 October 2017.)

The standard works through serial number 47312482. The weighted values are 32, 42, 12, 2, 6, 20, 72 and 14, which sum to 200. Dividing 200 by 11 gives 18 remainder 2, and 11 minus 2 is 9, so the complete self-checking number is 473124829. A number whose eleventh digit fails that test has almost always been mistyped or misread, which is the cheapest explanation to rule out before assuming a carrier problem.

Letter pairs that can never appear in a valid international number

Five whole ranges, 130 combinations in total, are reserved and cannot be assigned as service indicator values: JA to JZ, KA to KZ, SA to SZ, TA to TZ and WA to WZ. The standard blocks them to reduce the risk of confusion between S10 barcodes and barcodes that use ISO/IEC 15418 data identifiers, which occupy those same letters.

The practical use is diagnostic. A 2-9-2 string beginning with K or S is not an international postal item identifier, whatever the label around it suggests, and a search on a postal network will not resolve it. Combinations outside both the assigned and the reserved lists are unassigned rather than invalid, and may be brought into use by the Standards Board, so an unfamiliar prefix is not automatically a fake.

How long a tracking number stays unique before it can be reused

An S10 identifier may not be reused within 12 calendar months, and the standard recommends a gap of 24 calendar months or longer. Uniqueness is managed per country code by the single organisation that the UPU member state assigns, across every operator working under that state's authority, and operators may delegate serial ranges to individual customers only if the same 12-month floor is preserved.

Two consequences follow. An old number can eventually resolve to a different consignment, so a code kept from a delivery two years earlier is not a reliable handle. And because only one S10 identifier is ever assigned to a single item, a parcel showing two different 2-9-2 barcodes has picked up a second label somewhere in the chain rather than been split into two shipments.

The last two letters name the issuing authority, not the place of posting

The final two characters are the ISO 3166-1 code of the UPU member country under whose authority the identifier was issued, which is not the same thing as the country the item was handed over in. Operators that run extraterritorial offices of exchange issue identifiers carrying their own country code from premises abroad, and the standard names Swiss Post as an example, using the code CH for items despatched from its offices in other countries.

"This implies that the country code at the end of an S10 identifier cannot be used as a reliable indicator of the geographic origin of an item." (Universal Postal Union, standard S10-12, clause 6, approved 17 October 2017.)

A second limit sits alongside it. Where more than one operator works under a single ISO 3166-1 code, the country code cannot identify which of them issued the number, and some member states hold several codes for separate territories with a different designated operator behind each. The suffix narrows the field to one issuing authority in one code area, and that is the whole of what it proves.

Prefixes that are not S10 at all

Integrators and domestic parcel networks number their consignments outside the S10 scheme, so their prefixes decode against company documentation rather than against the UPU standard. UPS shipment numbers open with 1Z. FedEx tracking numbers are 12 digits with no letters, the same value as the air waybill number. DHL Express waybill numbers are 10 digits, and DHL Express and DHL eCommerce run separate systems with different number formats.

Domestic postal barcodes are a third family again. The USPS Intelligent Mail package barcode is a variable-length data string that follows the GS1-128 specification and uses application identifier 420 for the routing element, which is why a domestic USPS number is a long run of digits rather than a 2-9-2 string.

"An IMpb is the USPS-developed barcode that can be read by automated parcel-processing equipment and scanning devices and consists of a data string that generally follows the GS1-128 specification." (United States Postal Service, Domestic Mail Manual 204, clause 2.1.1.)

Operators frequently run both schemes side by side. Canada Post lists item numbers of 11 or 13 alphanumeric characters ending in CA alongside 16-digit numbers, so the presence or absence of letters signals which internal system generated the label.

What a prefix says about a marketplace order

A 13-character number ending in CN was issued under the authority of China's designated operator, and says nothing about which marketplace sold the goods. Orders from the same seller can arrive under an L range indicator on one shipment and a C range indicator on the next, because the range follows the postal product the consolidator bought rather than the storefront. Cross-border consignments routed through China Post and its partners are the largest source of S10 numbers most shoppers ever see.

The U range is the one worth recognising here. It covers letter-post items containing goods that need an identifier for supply-chain visibility, for example in an advance electronic data message, and explicitly without customer-oriented tracking. A U prefix therefore predicts a sparse event history by design, and the absence of scans is the intended behaviour of the product rather than a fault in it.

What the prefix cannot tell you about delivery

A service indicator describes the product class of an item at the moment the identifier was issued, and carries no information about position, customs status or delivery date. It does not change when the item moves, so the same 13 characters read identically on day 1 and on day 40. It does not encode weight, value, contents or the number of pieces in the consignment.

What it does establish is the entitlement behind the item, and that determines who can act. Registered, insured and EMS classes support the formal inquiry and claim process built on the CN 08 form, so an R, V or E prefix means an inquiry can be opened with the operator of origin. A U prefix, by contrast, was never sold as a tracked product. For reading the timeline itself rather than the prefix, the event wording matters more than the code, starting with what in transit means on a scan history.

What an S10 prefix gets confused with

Four different strings on a single parcel can look like the tracking number, and only one of them is the S10 identifier.

String	What it actually is	How to tell it apart
A 2-9-2 barcode that no network recognises	An S10-format barcode that is not an S10 identifier, applied by a customer or a handler for internal use	Correct shape, but the data content does not comply with the standard; the check digit often fails
A barcode added by the destination operator	A domestic identifier applied on arrival, sometimes with data identical to the original	Any S10-format barcode differing from the original must be obliterated before onward forwarding
DT followed by 12 numbers	A FedEx door tag left after a failed delivery attempt	Carries the DT prefix and is tied to an attempt, not to the consignment record
A 6 to 12 character order or reference number	A seller's internal order reference	Assigned by the shop before despatch, with no service indicator and no country code

The customs paperwork is a further source of confusion, because CN 22 and CN 23 are form numbers rather than item identifiers, and an S10 identifier is frequently printed on those forms as well as on the label.

Reading the prefix correctly answers one question and one only: what class of postal product the item was accepted as. Everything after that comes from the scan history, and the wider anatomy of a shipment reference is covered in the guide to the logistics tracking number.