This type of tunnel mechanism has been one of the first to be developed and has since then mostly been replaced by more sophisticated mechanisms. It uses IPv4-compatible IPv6 addresses on the tunnel endpoints. The address of the recipient node is specified by the packet that is being encapsulated. This method can only be used on router-to-host and host-to-host communication since these are the only schemes where one tunnel endpoint is also the recipient. Due to the use of particular addresses it only works on IPv6 over IPv4 tunnelling and not vice versa.

Such automatic tunnelling has now been deprecated, and has been written out of the [NG05] update. One reason for that lies in the ad-hoc nature of connectivity that results, lacking structure in the IPv6 domain; solutions such as ISATAP or 6to4 (see below) are generally considered preferable. The authors of this book strongly advise not to use this technique anymore, even where implementations still exist.