It is hypothesised that common toxins produced by bacteria growing in the respiratory tract following a viral infection are a cause of SIDS. This hypothesis is consistent with evidence that viral infections pre-dispose to SIDS, minimal morphological change at autopsy, maximum incidence during sleep and the age incidence of this disease. We present evidence of nasopharyngeal bacterial overgrowth in victims of SIDS and have developed a mathematical model based on the hypothesis which closely predicts the age distribution. The model predicts other age distribution patterns for less common toxins and these may apply to other diseases of childhood. The hypothesis can be tested and if sustained would offer hope of prevention.