Hi Romana - If you do R, why don't you try my package MCMC.qpcr (there is a detailed tutorial on my RG profile page...). It will work with 6 samples, although the power might be limited (you can compensate by analyzing more genes, though)

Hi Mikhail

Thank you for your suggestion, but I do not R :-( but I will download your package and maybe I could learn something from it.

HI,

It will work with as many samples you give it, however, as with everything the fewer the samples the less you can say for certain your samples are representative of a larger set.  If that is all you have though and those are the samples you will be analysing then Genorm will identify which are best reference genes to use

I agree with Matt