Ok, im fairly new to all of this, and I have been looking through the wiki for an answer to my question. If you had V0 and 256 CBR which would be better (ex have the best quality/size ratio)

thanks

If the rating-criteria is efficiency (quality/filesize ratio), then the rule of thumb always is "VBR", unless the two files vary drastically in average bitrate. And: The efficiency sweet spot for VBR, tends to be around V4-V3.

256 CBR has a significantly higher probability of audible artifacts, as there might be frames that "require" the full 320kbps to satisfy the psy model.

In either case, it is very unlikely that you will hear any artifacts in either V0 or 256 CBR files.

I think the most important thing is a clear understanding of what you want in the first place.

a) if you don't like the idea of wasting bits you're best off using VBR. As Lyx said efficiency is best with VBR. As far as quality is concerned if you don't care much about rare problems a setting like -V3 is fine, and if you want some security margin it's fine going higher up to -V0.

b) if you don't care about file size but want the best quality even for rare problems (though you usually won't get absolute perfection in these cases) and are willing to accept the fact that the quality setting is overkill most of the time using a setting like CBR 256 is a good choice.

First remark: there's a widespread misunderstanding about CBR: CBR internally does use a variable audio data bitrate though it's restricted (the bit reservoir concept). Due to this fact CBR 256 can locally produce a higher audio data bitrate than CBR 320, which is not just theory because bit reservoir usage is more restricted with 320 kbps frames with several encoders. So benski's reply is not totally correct.

Second remark: As you want to use Lame (a good choice) Lame has a specific ABR mode. ABR uses variable frame sizes and as such is a vbr method, but the internals are like with CBR with the important difference that there are no audio data bitrate restrictions like with CBR: when the bit reservoir's restrictions towards audio data bitrate variation are too strong ABR chooses a larger frame size.
So in a sense ABR is the best of both worlds when it's more about quality than about efficiency: bitrate variation is more conservative than with 'true' VBR (which depends totally on the psy model which can be weak on rare occasion), but audio data bitrate variation is not restricted as is done when using CBR.
It's common opinion (though not proven) that ABR has the better quality for low mp3 bitrates like 100 kbps and below. My personal experience is that ABR is also best in the high bitrate region (and be it only because the best VBR quality setting is -V0 which leads to an average bitrate of ~240 kbps with mixed-genre music).
My personal preference is using ABR 270.

All frames have the same buffer size restriction, be it 32 or 320 kbps. This size restriction limits the possible use of the bitreservoir. The ISO documents are not so clear in the point, how large an audio buffer may become. There are lots of different interpretations possible. In practice, all Windows systems come with a very restricted decoder made by FhG.

You certainly know better than I do. So 'CBR 256 can locally produce a higher audio data bitrate than CBR 320' is wrong and should be restated as 'Due to these restrictions 256 kbps frames can locally yield the same audio data bitrate like 320 kbps frames'.