Models of dream analysis either assume a continuum of waking and dreaming or the existence of two dissociated realities. Both approaches rely on different methodology. Whereas continuity models are based on content analysis, discontinuity models use a structural approach. In our study, we applied both methods to test specific hypotheses about continuity or discontinuity. We contrasted dream reports of congenitally deaf-mute and congenitally paraplegic individuals with those of non-handicapped controls. Continuity theory would predict that either the deficit itself or compensatory (...) experiences would surface in the dream narrative. We found that dream form and content of sensorially limited persons was indifferent from those of non-handicapped controls. Surprisingly, perceptual representations, even of modalities not experienced during waking, were quite common in the dream reports of our handicapped subjects. Results are discussed with respect to feedforward mechanisms and protoconsciousness theory of dreaming. (shrink)