If your density "standard" is the Proctor, then dry density should be done (noting; however, that determination of an accurate moisture content for cementitious materials is difficult....if you use the same method for moisture determination in the lab and the field you should be ok...as the actual moisture content becomes less relevant than the comparative moisture contents)

For the first time ever, I disagree with Ron . . . you should be using as a standard a percentage of the theoretical maximum air-free density. You might refer to: http://www.cement.org/pavements/pv_rcc_faq_density... . This is what our RCC spec said on my last project.

Hold on BigH...I'm not so sure we are disagreeing!
I agree that proper evaluation requires the determination of TAFD; however, this can be correlated for ease of testing to the more common direct comparison using common density methods and the dry density of the material. In either case, you will need to know the moisture content of the material (which, again, for cementitious materials can be a bit tricky).

All voids parameters need to be determined. The key is to reduce the voids in the mix to as low a point as practicable for the compaction.