We evaluate the demand for long term care (LTC) insurance prospects in a stated preference context, by means of the results of a choice experiment carried out on a representative sample of the Emilia-Romagna population. Choice modelling techniques have not been used yet for studying the demand for LTC services. In this paper these methods are first of all used in order to assess the relative importance of the characteristics which define some hypothetical insurance programmes and to elicit the willingness to pay for some LTC coverage prospects. Moreover, thanks to the application of a nested logit specification with 'partial degeneracy', we are able to model the determinants of the preference for status quo situations where no systematic cover for LTC exists. On the basis of this empirical model, we test for the effects of a series of socio-demographic variables as well as personal and household health state indicators.