Typically it's done via the latter--i.e., creating separate que files for the different conditions but it can work in many different ways. You can also create a couple questionnaires that are constant between conditions and then a set that contain the unique contributions of each condition. You can also use the same variable names in que files from different conditions so that the data end up in the same columns in the data file. Check out the overview section in the manual and also the section on data. That should give you some guidance in making your choices.