RFC 2554 PROPOSED STANDARD EMAIL · SMTP · AUTHENTICATION

SMTP AUTH: SMTP Service Extension for Authentication

Letting mail clients prove their identity to the mail server before sending mail.

On rfc-editor.org ↗
PUBLISHED
March 1999
AUTHORS
J. Myers
STREAM
Legacy
OBSOLETED BY
RFC 4954 — SMTP AUTH
COMMAND NAME
AUTH
SUCCESS REPLY CODE
235
CHALLENGE REPLY CODE
334
UNSUPPORTED MECHANISM CODE
504
AUTHENTICATION FAILURE CODE
535
DOI
10.17487/RFC2554

RFC text

Contents
Original text · RFC Editor View on rfc-editor.org ↗
Network Working Group                                           J. Myers
Request for Comments: 2554                       Netscape Communications
Category: Standards Track                                     March 1999


                         SMTP Service Extension
                           for Authentication

Status of this Memo

   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice

   Copyright (C) The Internet Society (1999).  All Rights Reserved.


1. Introduction

   This document defines an SMTP service extension [ESMTP] whereby an
   SMTP client may indicate an authentication mechanism to the server,
   perform an authentication protocol exchange, and optionally negotiate
   a security layer for subsequent protocol interactions.  This
   extension is a profile of the Simple Authentication and Security
   Layer [SASL].


2. Conventions Used in this Document

   In examples, "C:" and "S:" indicate lines sent by the client and
   server respectively.

   The key words "MUST", "MUST NOT", "SHOULD", "SHOULD NOT", and "MAY"
   in this document are to be interpreted as defined in "Key words for
   use in RFCs to Indicate Requirement Levels" [KEYWORDS].


3. The Authentication service extension


   (1) the name of the SMTP service extension is "Authentication"

   (2) the EHLO keyword value associated with this extension is "AUTH"
   (3) The AUTH EHLO keyword contains as a parameter a space separated
       list of the names of supported SASL mechanisms.

   (4) a new SMTP verb "AUTH" is defined

   (5) an optional parameter using the keyword "AUTH" is added to the
       MAIL FROM command, and extends the maximum line length of the
       MAIL FROM command by 500 characters.

   (6) this extension is appropriate for the submission protocol
       [SUBMIT].


4. The AUTH command

   AUTH mechanism [initial-response]

     Arguments:
         a string identifying a SASL authentication mechanism.
         an optional base64-encoded response

     Restrictions:
         After an AUTH command has successfully completed, no more AUTH
         commands may be issued in the same session.  After a successful
         AUTH command completes, a server MUST reject any further AUTH
         commands with a 503 reply.

         The AUTH command is not permitted during a mail transaction.

     Discussion:
         The AUTH command indicates an authentication mechanism to the
         server.  If the server supports the requested authentication
         mechanism, it performs an authentication protocol exchange to
         authenticate and identify the user.  Optionally, it also
         negotiates a security layer for subsequent protocol
         interactions.  If the requested authentication mechanism is not
         supported, the server rejects the AUTH command with a 504
         reply.

         The authentication protocol exchange consists of a series of
         server challenges and client answers that are specific to the
         authentication mechanism.  A server challenge, otherwise known
         as a ready response, is a 334 reply with the text part
         containing a BASE64 encoded string.  The client answer consists
         of a line containing a BASE64 encoded string.  If the client
         wishes to cancel an authentication exchange, it issues a line
         with a single "*".  If the server receives such an answer, it
         MUST reject the AUTH command by sending a 501 reply.

         The optional initial-response argument to the AUTH command is
         used to save a round trip when using authentication mechanisms
         that are defined to send no data in the initial challenge.
         When the initial-response argument is used with such a
         mechanism, the initial empty challenge is not sent to the
         client and the server uses the data in the initial-response
         argument as if it were sent in response to the empty challenge.
         Unlike a zero-length client answer to a 334 reply, a zero-
         length initial response is sent as a single equals sign ("=").
         If the client uses an initial-response argument to the AUTH
         command with a mechanism that sends data in the initial
         challenge, the server rejects the AUTH command with a 535
         reply.

         If the server cannot BASE64 decode the argument, it rejects the
         AUTH command with a 501 reply.  If the server rejects the
         authentication data, it SHOULD reject the AUTH command with a
         535 reply unless a more specific error code, such as one listed
         in section 6, is appropriate.  Should the client successfully
         complete the authentication exchange, the SMTP server issues a
         235 reply.

         The service name specified by this protocol's profile of SASL
         is "smtp".

         If a security layer is negotiated through the SASL
         authentication exchange, it takes effect immediately following
         the CRLF that concludes the authentication exchange for the
         client, and the CRLF of the success reply for the server.  Upon
         a security layer's taking effect, the SMTP protocol is reset to
         the initial state (the state in SMTP after a server issues a
         220 service ready greeting).  The server MUST discard any
         knowledge obtained from the client, such as the argument to the
         EHLO command, which was not obtained from the SASL negotiation
         itself.  The client MUST discard any knowledge obtained from
         the server, such as the list of SMTP service extensions, which
         was not obtained from the SASL negotiation itself (with the
         exception that a client MAY compare the list of advertised SASL
         mechanisms before and after authentication in order to detect
         an active down-negotiation attack).  The client SHOULD send an
         EHLO command as the first command after a successful SASL
         negotiation which results in the enabling of a security layer.

         The server is not required to support any particular
         authentication mechanism, nor are authentication mechanisms
         required to support any security layers.  If an AUTH command
         fails, the client may try another authentication mechanism by
         issuing another AUTH command.

         If an AUTH command fails, the server MUST behave the same as if
         the client had not issued the AUTH command.

         The BASE64 string may in general be arbitrarily long.  Clients
         and servers MUST be able to support challenges and responses
         that are as long as are generated by the authentication
         mechanisms they support, independent of any line length
         limitations the client or server may have in other parts of its
         protocol implementation.

     Examples:
         S: 220 smtp.example.com ESMTP server ready
         C: EHLO jgm.example.com
         S: 250-smtp.example.com
         S: 250 AUTH CRAM-MD5 DIGEST-MD5
         C: AUTH FOOBAR
         S: 504 Unrecognized authentication type.
         C: AUTH CRAM-MD5
         S: 334
         PENCeUxFREJoU0NnbmhNWitOMjNGNndAZWx3b29kLmlubm9zb2Z0LmNvbT4=
         C: ZnJlZCA5ZTk1YWVlMDljNDBhZjJiODRhMGMyYjNiYmFlNzg2ZQ==
         S: 235 Authentication successful.



5. The AUTH parameter to the MAIL FROM command

   AUTH=addr-spec

   Arguments:
       An addr-spec containing the identity which submitted the message
       to the delivery system, or the two character sequence "<>"
       indicating such an identity is unknown or insufficiently
       authenticated.  To comply with the restrictions imposed on ESMTP
       parameters, the addr-spec is encoded inside an xtext.  The syntax
       of an xtext is described in section 5 of [ESMTP-DSN].

   Discussion:
       The optional AUTH parameter to the MAIL FROM command allows
       cooperating agents in a trusted environment to communicate the
       authentication of individual messages.

       If the server trusts the authenticated identity of the client to
       assert that the message was originally submitted by the supplied
       addr-spec, then the server SHOULD supply the same addr-spec in an
       AUTH parameter when relaying the message to any server which
       supports the AUTH extension.

       A MAIL FROM parameter of AUTH=<> indicates that the original
       submitter of the message is not known.  The server MUST NOT treat
       the message as having been originally submitted by the client.

       If the AUTH parameter to the MAIL FROM is not supplied, the
       client has authenticated, and the server believes the message is
       an original submission by the client, the server MAY supply the
       client's identity in the addr-spec in an AUTH parameter when
       relaying the message to any server which supports the AUTH
       extension.

       If the server does not sufficiently trust the authenticated
       identity of the client, or if the client is not authenticated,
       then the server MUST behave as if the AUTH=<> parameter was
       supplied.  The server MAY, however, write the value of the AUTH
       parameter to a log file.

       If an AUTH=<> parameter was supplied, either explicitly or due to
       the requirement in the previous paragraph, then the server MUST
       supply the AUTH=<> parameter when relaying the message to any
       server which it has authenticated to using the AUTH extension.

       A server MAY treat expansion of a mailing list as a new
       submission, setting the AUTH parameter to the mailing list
       address or mailing list administration address when relaying the
       message to list subscribers.

       It is conforming for an implementation to be hard-coded to treat
       all clients as being insufficiently trusted.  In that case, the
       implementation does nothing more than parse and discard
       syntactically valid AUTH parameters to the MAIL FROM command and
       supply AUTH=<> parameters to any servers to which it
       authenticates using the AUTH extension.

   Examples:
       C: MAIL FROM:<e=mc2@example.com> AUTH=e+3Dmc2@example.com
       S: 250 OK

6. Error Codes

   The following error codes may be used to indicate various conditions
   as described.

   432 A password transition is needed

   This response to the AUTH command indicates that the user needs to
   transition to the selected authentication mechanism.  This typically
   done by authenticating once using the PLAIN authentication mechanism.

   534 Authentication mechanism is too weak

   This response to the AUTH command indicates that the selected
   authentication mechanism is weaker than server policy permits for
   that user.

   538 Encryption required for requested authentication mechanism

   This response to the AUTH command indicates that the selected
   authentication mechanism may only be used when the underlying SMTP
   connection is encrypted.

   454 Temporary authentication failure

   This response to the AUTH command indicates that the authentication
   failed due to a temporary server failure.

   530 Authentication required

   This response may be returned by any command other than AUTH, EHLO,
   HELO, NOOP, RSET, or QUIT.  It indicates that server policy requires
   authentication in order to perform the requested action.

7. Formal Syntax

   The following syntax specification uses the augmented Backus-Naur
   Form (BNF) notation as specified in [ABNF].

   Except as noted otherwise, all alphabetic characters are case-
   insensitive.  The use of upper or lower case characters to define
   token strings is for editorial clarity only.  Implementations MUST
   accept these strings in a case-insensitive fashion.

   UPALPHA         = %x41-5A            ;; Uppercase: A-Z

   LOALPHA         = %x61-7A            ;; Lowercase: a-z

   ALPHA           = UPALPHA / LOALPHA  ;; case insensitive

   DIGIT           = %x30-39            ;; Digits 0-9

   HEXDIGIT        = %x41-46 / DIGIT    ;; hexidecimal digit (uppercase)

   hexchar         = "+" HEXDIGIT HEXDIGIT

   xchar           = %x21-2A / %x2C-3C / %x3E-7E
                     ;; US-ASCII except for "+", "=", SPACE and CTL

   xtext           = *(xchar / hexchar)

   AUTH_CHAR       = ALPHA / DIGIT / "-" / "_"

   auth_type       = 1*20AUTH_CHAR

   auth_command    = "AUTH" SPACE auth_type [SPACE (base64 / "=")]
                     *(CRLF [base64]) CRLF

   auth_param      = "AUTH=" xtext
                       ;; The decoded form of the xtext MUST be either
                       ;; an addr-spec or the two characters "<>"

   base64          = base64_terminal /
                     ( 1*(4base64_CHAR) [base64_terminal] )

   base64_char     = UPALPHA / LOALPHA / DIGIT / "+" / "/"
                     ;; Case-sensitive

   base64_terminal = (2base64_char "==") / (3base64_char "=")

   continue_req    = "334" SPACE [base64] CRLF
   CR              = %x0C           ;; ASCII CR, carriage return

   CRLF            = CR LF

   CTL             = %x00-1F / %x7F ;; any ASCII control character and DEL

   LF              = %x0A           ;; ASCII LF, line feed

   SPACE           = %x20           ;; ASCII SP, space




8. References

   [ABNF]      Crocker, D. and P. Overell, "Augmented BNF for Syntax
               Specifications: ABNF", RFC 2234, November 1997.

   [CRAM-MD5]  Klensin, J., Catoe, R. and P. Krumviede, "IMAP/POP
               AUTHorize Extension for Simple Challenge/Response", RFC
               2195, September 1997.

   [ESMTP]     Klensin, J., Freed, N., Rose, M., Stefferud, E. and D.
               Crocker, "SMTP Service Extensions", RFC 1869, November
               1995.

   [ESMTP-DSN] Moore, K, "SMTP Service Extension for Delivery Status
               Notifications", RFC 1891, January 1996.

   [KEYWORDS]  Bradner, S., "Key words for use in RFCs to Indicate
               Requirement Levels", BCP 14, RFC 2119, March 1997.

   [SASL]      Myers, J., "Simple Authentication and Security Layer
               (SASL)", RFC 2222, October 1997.

   [SUBMIT]    Gellens, R. and J. Klensin, "Message Submission", RFC
               2476, December 1998.

   [RFC821]    Postel, J., "Simple Mail Transfer Protocol", STD 10, RFC
               821, August 1982.

   [RFC822]    Crocker, D., "Standard for the Format of ARPA Internet
               Text Messages", STD 11, RFC 822, August 1982.

9. Security Considerations

   Security issues are discussed throughout this memo.

   If a client uses this extension to get an encrypted tunnel through an
   insecure network to a cooperating server, it needs to be configured
   to never send mail to that server when the connection is not mutually
   authenticated and encrypted.  Otherwise, an attacker could steal the
   client's mail by hijacking the SMTP connection and either pretending
   the server does not support the Authentication extension or causing
   all AUTH commands to fail.

   Before the SASL negotiation has begun, any protocol interactions are
   performed in the clear and may be modified by an active attacker.
   For this reason, clients and servers MUST discard any knowledge
   obtained prior to the start of the SASL negotiation upon completion
   of a SASL negotiation which results in a security layer.

   This mechanism does not protect the TCP port, so an active attacker
   may redirect a relay connection attempt to the submission port
   [SUBMIT].  The AUTH=<> parameter prevents such an attack from causing
   an relayed message without an envelope authentication to pick up the
   authentication of the relay client.

   A message submission client may require the user to authenticate
   whenever a suitable SASL mechanism is advertised.  Therefore, it may
   not be desirable for a submission server [SUBMIT] to advertise a SASL
   mechanism when use of that mechanism grants the client no benefits
   over anonymous submission.

   This extension is not intended to replace or be used instead of end-
   to-end message signature and encryption systems such as S/MIME or
   PGP.  This extension addresses a different problem than end-to-end
   systems; it has the following key differences:

      (1) it is generally useful only within a trusted enclave

      (2) it protects the entire envelope of a message, not just the
          message's body.

      (3) it authenticates the message submission, not authorship of the
          message content

      (4) it can give the sender some assurance the message was
          delivered to the next hop in the case where the sender
          mutually authenticates with the next hop and negotiates an
          appropriate security layer.

   Additional security considerations are mentioned in the SASL
   specification [SASL].



10. Author's Address

   John Gardiner Myers
   Netscape Communications
   501 East Middlefield Road
   Mail Stop MV-029
   Mountain View, CA 94043

   EMail: jgmyers@netscape.com

11.  Full Copyright Statement

   Copyright (C) The Internet Society (1999).  All Rights Reserved.

   This document and translations of it may be copied and furnished to
   others, and derivative works that comment on or otherwise explain it
   or assist in its implementation may be prepared, copied, published
   and distributed, in whole or in part, without restriction of any
   kind, provided that the above copyright notice and this paragraph are
   included on all such copies and derivative works.  However, this
   document itself may not be modified in any way, such as by removing
   the copyright notice or references to the Internet Society or other
   Internet organizations, except as needed for the purpose of
   developing Internet standards in which case the procedures for
   copyrights defined in the Internet Standards process must be
   followed, or as required to translate it into languages other than
   English.

   The limited permissions granted above are perpetual and will not be
   revoked by the Internet Society or its successors or assigns.

   This document and the information contained herein is provided on an
   "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
   TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
   BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
   HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

AI ELI5

On this page
PROBLEMSMTP as originally defined allowed anyone to connect and send mail, with no way for a server to verify who is sending. Mail clients and submission agents needed a standard way to authenticate.
MECHANISMRFC 2554 adds an AUTH command to SMTP that lets a client indicate a SASL authentication mechanism, perform an authentication exchange, and optionally negotiate a security layer (encryption) for subsequent commands.
WHY IT WAS REPLACEDSMTP AUTH became the standard way for mail clients to authenticate to mail servers. It was superseded by RFC 4954, which clarified and extended the mechanism.

§1The Problem and the Solution

Simple Mail Transfer Protocol (SMTP) was designed to be a simple protocol for sending mail between systems on a trusted network. It includes no authentication: any client that connects can claim to be anyone and send mail on their behalf. This worked when the Internet was small and trusted, but it enabled spam and spoofing.

RFC 2554 adds an authentication mechanism to SMTP using the Simple Authentication and Security Layer (SASL), a framework for pluggable authentication mechanisms. A mail client can now connect to a server and prove its identity using a mechanism that both sides support—such as login with a password, a challenge-response hash, or even Kerberos.

The extension also allows negotiation of a security layer (typically TLS encryption) for the rest of the session, protecting the subsequent mail commands from snooping.

Before, anyone could pretend to be you when sending mail to a server. Now the server can ask 'who are you?' and the client can prove it by using a password or other proof.

§2The AUTH Command

A client initiates authentication with the AUTH command, specifying a SASL mechanism (such as PLAIN, LOGIN, CRAM-MD5, or DIGEST-MD5) and optionally including an initial response. The syntax is AUTH mechanism [initial-response].

If the server recognizes the mechanism, it sends a 334 challenge (a BASE64-encoded string). The client responds with a BASE64-encoded answer. The server may send more challenges; the client answers each one. If at any point the client wants to cancel, it sends a single asterisk. If the authentication succeeds, the server sends a 235 reply. If it fails, the server sends a 535 reply (or a more specific error code).

Once authentication succeeds, no further AUTH commands are allowed in that session. The authenticated identity is available for subsequent commands, such as the MAIL FROM command, which can declare on whose behalf the message is being sent.

The client says 'I want to use CRAM-MD5.' The server sends a challenge. The client sends an answer. If the answer is right, the server says 'OK, it's you.'

SMTP authentication exchange (illustrative, not from RFC)
ClientServer
  1. EHLO
  2. 250 AUTH PLAIN CRAM-MD5
  3. AUTH CRAM-MD5
  4. 334 [challenge]
  5. [response]
  6. 235 Authenticated

§3Choosing an Authentication Mechanism

After a client connects with EHLO (Extended HELO), the server advertises which SASL mechanisms it supports using the AUTH keyword in its response. The client then chooses one mechanism and issues an AUTH command with that mechanism name.

If the server does not support the requested mechanism, it replies with a 504 error. The client may then try a different mechanism by issuing another AUTH command, unless authentication has already succeeded.

The initial-response parameter allows optimization: some SASL mechanisms (like PLAIN) send no data in the server's first challenge, expecting the client to send all credentials in response to an empty challenge. By including the initial-response in the AUTH command itself, the client saves a round trip.

The server says 'I support CRAM-MD5 and PLAIN.' The client picks one and tries it. If the server doesn't support it, the client can try another.

§4Optional Encryption After Authentication

Some SASL mechanisms can negotiate a security layer—typically TLS-style encryption—that protects all subsequent commands from that point forward. After the successful 235 response, the security layer takes effect immediately, and the SMTP conversation resets to its initial state.

Both client and server must discard any state they learned before authentication (such as the list of SMTP extensions) because an attacker could have modified it. After the security layer is active, the client should send a new EHLO command to learn the server's capabilities over the encrypted channel.

This provides confidentiality and integrity for sensitive commands like MAIL FROM and message data, protecting them from network sniffing or tampering.

After proving who you are, you can also agree to lock the door so no one else can read your mail orders. The server forgets everything it heard before the lock, so an attacker can't trick it.

§5Tracking Message Origin in Relays

RFC 2554 adds an optional AUTH parameter to the MAIL FROM command. When a client sends a message on behalf of a user, it can include AUTH=user@example.com to assert who originally submitted the message.

If the server trusts the authenticated identity of the client, and the client is claiming the message came from a particular user, the server may include the same AUTH parameter when relaying the message to another server. This creates a chain of trust: downstream servers can verify that the message was authenticated at an earlier hop.

If the client is not authenticated, or if the server does not trust the client to assert message origin, the server uses AUTH=<> (empty brackets), meaning the original submitter is unknown or not authenticated. This prevents a malicious or compromised client from forging the appearance of an authenticated origin.

When a mail relay passes a message to another server, it can say 'the original sender proved they are alice@example.com.' Downstream servers can trust this if they trust the relay.

§6Error Codes and Security Considerations

RFC 2554 defines several error codes for specific authentication failures. A 432 reply means the user must first authenticate using the PLAIN mechanism (password transition). A 534 reply indicates the chosen mechanism is too weak for that user or server policy. A 538 reply means the mechanism requires encryption (TLS) to be enabled first.

Security is critical: before SASL negotiation begins, protocol interactions are unencrypted and can be modified by an attacker. Clients and servers must discard any unencrypted knowledge obtained before authentication, especially if a security layer is negotiated afterward. A client requiring secure submission should be configured to never send mail to a server unless the connection is both authenticated and encrypted.

The extension is not a replacement for end-to-end message signing or encryption (like S/MIME or PGP). It protects the transaction with the server but not the message content itself after delivery.

If something goes wrong during login—the password is weak, or the connection isn't encrypted enough—the server sends a specific error so the client knows what to fix.

§7References