Not many people use a normalization parameter for agonist induced responses. Once you have given the tissue time to settle in the myograph (typically an hour or so) you can use a bolus of KCl just to check that the vessel is responding normally to depolarization. Then wash this out several times over a 30 min period then construct your curve and just use absolute values. If there is a difference in KCl between the two groups and you normalize your curve against this then it will appear "normal" when in fact it isn't. Do you plan to look at endothelial function too? And are you ok with your drug solutions?

**solutions I meant, not solutions

***dilutions

I recognize the problem. It depends on what you expect. If the disease affects the EC50 of the stimulus, there is no problem so normalization is not needed. However, for Emax it is essential. Maybe, u can use KCl (as suggested before) or CaCl2contraction to normalize because those may not be affected by the disease in contrast to an receptor mediated contraction. If you measure in a blood vessel, 'just' cut the vessel rings accurately and put a well defined pressure on the rings. Measure many rings. For ileum reproducible pieces is not possible. Waiting for suggestions too.

Like Marijke said, if you cut the rings accurately each time and use same pretension or pressure then your results will be fine. n=5 or more is the norm for myograph data. If you use a bolus of KCl before you do you curve, washing out in between obviously, then at least you will have that info if you need it. It is normal in any case to use KCl initially. People generally say it "conditions" the tissue. 

If you are also studying endothelial function, e.g. with acetylcholine or bradykinin, don't do your curve at maximal agonist constriction; rather use a concentration of agonist that gives around 80% maximal constriction before doing your relaxation curve. 

you can use kcl upto 60mM , dont use more than that. even it depends on your aim of disease.