Tough question. Most radio mixes I hear have more compression that the album versions. Strangely, most pop radio stations squash the hell out of programme material anyways and EQ with a big 'smile' freq response. I'd be tempted to leave more dynamics in there and counteract the smile EQ with a 'frown' and see what happens. If this is done right there is a small chance that the tune might some comparable to how you intended it to sound when it gets transmitted. However, if it goes wrong it will sound awful.

In my experince, radio station engineers don't have a clue about much of anything, let alone sound.

volume shouldn't be an issue, as the radio station will crush the crap out of anyways....as far as eq, i don't know that i'd eq any differently..

check out my stuff at
http://www.mp3.com/thedefinition

"How do I master to make my shit as "loud" as every piece of shit on the radio?" I DON"T!!!!
I master appropriate to the music and the context. YOU REMEMBER MUSIC DON'T YOU? IT'S WHY WE EXIST!!!!!!

I figured out a cheap way to solve the problem of getting the volume of each song the same. I have a stereo reciever with meters on it. It has a calibrated volume knob and I simply set it the same each time and play the song through it from my computer. If you are using a stand alone recorder or a cassette deck or whatever, you can also play it back through the mixer and watch the level meters. At least you have a good guide as to the volume levels.

The biggest difficulty is song to song level within your project and these tricks help solve that problem too.

The very least you could/should do is watch the (LED) meters on your mixer or the peak meters on your software program you are mastering to/with & adjust levels to match.
I use a pair of VUs driven by a small power amp that receives audio signals from a mixer (used as a patch bay). I have the mixer set to send audio to the VUs at the same level from all sources so that I can easily compare levels.
Also, some programs have a way to calculate the level of a song file so you will know exactly where your's stands.

Boom! How about you go out and get yourself the newest version of T-Racks. Keep an eye on your levels should probably be the main concern right. I was told to save the mix down a few decibels below zero so that you'll have room to boost when you master. I use either the newest version of wavelab or the newest version of t racks. The new t racks comes complete with several mastering settings and a a meter to keep track of your leveling. 0 dbs is 0 decibels no matter what type track, correct? If i'm wrong let me know but i've come with excellent sounding mixes and masters using T racks and wavelab.

Keeping an eye on your peak levels may give you some idea of how loud each track is relative to each. But, since the 'loudness' of your songs is actually related to the RMS (average) level, this could be misleading. And since your ears perceive different frequencies at different levels even if they are in fact the same, you could fool yourself too. The best way to match levels is to listen to a reference CD (one that you think sounds loud enough) and then to your track and go back and forth, making adjustments to the amount of compression, etc. If you get it to where you can't tell the difference, no one else will either after they run it through a gazillion compressors and broadcast it.

The RMS level as a guide is a good start. I compile and assemble and master a monthly compilation CD for a magazine in the Atlanta area. The material ranges from recordings done in professional studios to project studio digital recordings done in a basement. For each cut I will always do a quick check using Sound Forge's "statistics" function. This will give me the average RMS level as well as peak level information. This gives me a starting point to see which ones may need to be louder and which may need to come down in level.
Generally speaking, I have found when the RMS values are within a db or two of each other the songs will sound more or less equal in volume. This is only a very rudimentary starting point though.
The bottom line, as was stated before, is you have to listen and adjust accordingly.

Chuck

Run a c.d. player(in stereo) through two channels of your mixer(no E.Q.) with a song of a group that has the same style or dynanics you're going for(also, no E.Q.). Then, run your new songs' master out through a normalizer and A/B(compare)the two stereo outs till you get as close as you can(level-wise). If you do this, when your song gets played on the radio, it will sound like the pre-recorded groups disc does on the radio. Just remember to test out and A/B your recording and the other disc in your car and boom box so that you can fine tune if necessary. Unfortunately, it won't sound like it does on the radio through your system cuz' you they do compress the hell out of it.

Make sure that nothing is clipping in your master. Clipping will cause an FM station's transmitter processing to turn your song down.