Technically, yes, this concerns English pronunciation.  But I'm not
interested at this time in idiodialectic variation.  I'm just looking
for a good list of English phonemes, regardless of how they're actually
realized in 'lect X.

A dictionary pronunciation guide seemed like a good place to start,
so I grabbed the one from  But I'd be interested to hear
comments from you fine folks.

Here's the M-W guide:

M-W                                     X-SAMPA
\&\ as a and u in abut                  /V/ or /@/
\a\ as a in ash                         /{/
\A\ as a in ace                         /e/
\\ as o in mop                         /A/
\au\ as ou in out                       /aU/
\ch\ as ch in chin                      /t_S/
\e\ as e in bet                         /E/
\E\ as ea in easy                       /i/
\g\ as g in go                          /g/
\i\ as i in hit                         /I/
\I\ as i in ice                         /aj/
\j\ as j in job                         /d_Z/
\[ng]\ as ng in sing                    /N/
\O\ as o in go                          /o/
\o\ as aw in law                        /O/
\oi\ as oy in boy                       /oj/
\th\ as th in thin                      /T/
\<ul>th</ul>\ as th in the              /D/
\\ as oo in loot                       /u/
\u\ as oo in foot                       /U/
\y\ as y in yet                         /j/
\zh\ as si in vision                    /Z/

It also includes these symbols:

        \<sup>&</sup>\ as e in kitten
        \&r\as ur/er in further

Both of the above are really just examples of syllabicizing the
following consonant, though, so I don't consider them phonemes.

There are, of course, many other phones in English; they were left out of the
guide because the dictionary uses standard English orthography which is in
these cases unambiguous.  Here they are, as found in actual
dictionary entries:

\b\     /b/
\d\     /d/
\f\     /f/
\h\     /h/
\k\     /k/
\l\     /l/
\m\     /m/
\n\     /n/
\p\     /p/
\r\     /r\/
\s\     /s/
\sh\    /S/
\t\     /t/
\v\     /v/
\w\     /w/
\hw\    /W/
\z\     /z/

So by my count that's 39 phonemes.

Additions, deletions, modifications?  Again, I'm interested
in identity, not phonetics; for instance, the 'o' in 'mop' is
a separate phoneme from the one in 'mope' regardless of how you
personally make the distinction, and my choice of /O/ and /o/ for
their respective X-SAMPA representations is not intended to reflect
anyone's actual pronunciation (not even mine, since my long o definitely
has an off-glide).  In general, wherever appropriate distinct
qualitative symbols were available, I used them in lieu of using
distinctions of length (e.g. /e/ vs /e:/) or diphthongalization
(e.g. /e/ vs /ej/).