Views on the interpretation of the protective housing issue have until now been brought out in a scarcely organized manner. In this study the opinions on this matter are listed in a systematic way. It turns out that protective housing needs to have a receptive, a rehabilitation and a residence function; each PH-unit is focussed on one of these three functions. Associating this with the desired differences in the size of a group, twelve variations of PH can be distinguished. These variations have to be tested on their empirical value in a follow-up study. Within these margins the offered support needs to be flexible in nature. Research data have among other things shown that the admission of new target groups may be avoided by having temporarily the first admission to PH only after a stay in a General Psychiatric Hospital. Budgettary neutrality seems to be a hardly realistic demand.