Updated 2018 May 25
You can just click here for the list, and/or here for a list with links to maps for each site. I also have a list of corrections and additions to the MPC's list. I'd suggest reading the following explanations, though.
The MPC provides a list of MPC observatory codes. For each, you get the three-character code, longitude, parallax constants rho cos phi' and rho sin phi' (ρsin(φ') and ρcos(φ')) and the observatory name.
My list adds the geodetic latitudes and altitudes (in meters) of these observatory codes. (The geodetic latitude is what people usually mean when they say "latitude".) A few example lines follow :
Code Longitude Latitude Altitude ρcos(φ') ρsin(φ') region Obs Name ---- --------- ----------- ------- --------- ---------- --------- ---------------- 000 0.0000 +51.477379 65.985 0.62411 +0.77873 UK Greenwich 001 0.1542 +51.051835 267.895 0.62992 +0.77411 UK Crowborough 002 0.62 +51.652979 3041.392 0.622 +0.781 UK Rayleigh 003 3.90 +43.650525 2466.725 0.725 +0.687 France Montpellier 004 1.4625 +43.612279 177.272 0.72520 +0.68627 France Toulouse 005 2.23100 +48.805069 157.735 0.65989 +0.74887 France Meudon
I've also added some geographic region information so you can tell in which country a given MPC station is, and added a few spaces to allow for additional precision and for the likelihood that MPC codes will be expanded eventually to four characters. (The MPC file allows for a precision of five digits in longitude, or about one meter; but you only get six digits in the parallax constants, or about 6.3 meter precision. And plenty of stations have fewer digits than that; in those cases, the altitude is omitted, which is preferable to giving you a value that's basically garbage. In all cases, the number of digits shown for longitude and parallax constants matches the original list from the MPC.)
Note that the altitudes are approximate to begin with, just because of the six-digit parallax constants; but in addition, they are ellipsoidal heights, not geoid ("above mean sea level") heights. That can introduce a couple of dozen meters of difference. Before parallax constants are computed, one really should ensure that the altitude is above the ellipsoid, but I've no idea how consistently that has been done. (It makes minimal difference for asteroid astrometry, but a couple of dozen meters might be significant in other contexts. For example, people observing GPS satellites to determine timing errors may be able to detect the difference.) We definitely know that the geoid height correction was not correctly included for Mauna Kea and for (J95) Great Sheffield. (Though Mauna Kea observatory positions have many issues beyond just ignoring the geoid height, mostly involving telescopes scattered over half a kilometer being given only one observatory code.)
On top of that, I've had doubts that people actually provide their coordinates to six-meter precision, or even sixty-meter precision; that's probably especially true for older observatories measured in the pre-GPS era.
The source code used to create this list is available. It reads in the MPC file, computes latitudes and altitudes, figures out the region for each observatory, and adds spaces.
I also have a list of some additional observatories that may be useful. Among other things, it gives precise positions for all telescopes on Mauna Kea, Cerro Tololo, Cerro Paranal, La Silla, Haleakalā, Kitt Peak, and for the DSN stations.
Note on geodetic latitude: The geodetic latitude is (close to) the angle between your local vertical and the equatorial plane. The real angle is subject to a lot of local variations; for example, if you're next to a mountain, it'll pull the local vertical away from what you'd otherwise expect, deflecting both latitude and longitude. So instead, we model the earth as being a slightly flattened ellipsoid, and take the local vertical to that. This idealized latitude is what you normally see on maps and GPS. Wikipedia has a decent explanation of this.
Note on parallax constants: The parallax constants ρsin(φ') and ρcos(φ') provide cylindrical coordinates: ρcos(φ') tells you how far the observatory is from the earth's axis, and ρsin(φ') tells you how far above or below the earth's plane it is, both in units of the earth's equatorial radius, 6378.140 km. Put them together with a longitude, and you've specified exactly where the observatory is. φ' is the geocentric latitude; ρ is the distance from the center of the earth.