Obligatory nasal breathing has been suggested in the past as a contributor to sudden infant death syndrome (SIDS): nasal obstruction would result in death as infants were unable to breathe orally. To test this hypothesis, we studied 55 normal and 14 near-miss for SIDS infants during a whole-night polysomnography. On several occasions, the infant nares were gently occluded by the fingertips of the investigator. Infants continued to make respiratory efforts against the occluded nose for a variable time (apnoea time), then opened the mouth and started to breathe through it. Mean apnoea time in normal infants was 4.76 +/- 3.41 s (means +/- SD), and 6.54 +/- 4.25 s in near-miss for SIDS ones. These figures were not significantly different. Analysis according to sleep stage (quiet sleep: 4.08 +/- 3.24 s in normals and 6.50 +/- 4.18 s in near-miss for SIDS ones; active sleep: 6.54 +/- 3.67 s in normals and 6.58 +/- 4.76 s in near-miss for SIDS ones) did not disclose any significant difference between groups. There was no significant relationship between apnoea time and age in either group. In many cases, an arousal preceded the resumption of (oral) flow. However, in almost half of the occlusions, oral breathing was initiated during continuing sleep. We conclude: 1) infants are not obligatory nasal breathers, and 2) the nasal obstruction hypothesis should be discarded in the etiology of SIDS.