How do I determine whether two dose-response curves are parallel?

Assuming your dose-response curves follow the typical sigmoidal shape, asking whether two curves is parallel is the same as asking whether their slope factors (Hill slopes) differ significantly.

These instructions are for Prism 5, but they also can be adapted for use with Prism 4.

1. Enter, or transform, your data so X is log(concentration) and Y is response.
2. Click "Analyze" and choose nonlinear regression.
3. Go to the equation section for Dose-response stimulation (or inhibition).
4. Choose 'log(agonist) vs. response (variable slope)"
5. Optional. Go to the Constrain tab, and constrain Top and/or  Bottom to constant values, or share those parameters between the curves.
6. Go to the compare tab.
   
7. Check the option "Do the best fit values of selected parameters differ between data sets", and check the option box for the Hill Slope.
8. Click OK.
9. Look at the P value for the comparison at the top of the results table. If the P value is low, then the slopes of the curves differ significantly from one another.