qPCR Ct's are doubled for every single increase. The difference between a Ct of 25 and 26 means that the sample at 25 has TWICE as much of the target. For data that I would believe, I would like the SD to be as little as possible. If I was doing technical replicates on a plate, I would not accept a difference of more than .5 Ct, but thats just a personal flavor. Whatever will get you to the P-value you seek is your answer.

To an extent Craig is correct. Generally speaking, you wouldn't accept SDs of 0.5. That said, for genes with very low expression you will run into something called the Poisson Distribution. Short version; with very few molecules in a tube, it is difficult to accurately separate them into multiple reaction tubes. If you do a dilution series, you will see a swing out to one side of the regression line into a curve shape and your variation will increase.

To give you an idea, a Ct of 40 corresponds to ~1 RNA molecule in the 10 (or 20) uL reaction. 39 would be ~2 RNA molecules. I don't know about you, but I wouldn't have much confidence accurately diluting the two molecules in a 39 Ct reaction to 2 40 Ct reactions. Just as likely to get one 39 and one no reading.

Good luck!

Hi Cody,

The answer is for sure as little as possible. Mine is always controlled less than 0.5.