The relationship between the dental occlusion and temporomandibular disorders (TMDs) has been one of the most controversial topics in the dental community. In a large epidemiological cross-sectional survey - the Study of Health in Pomerania (Germany) - associations between 15 occlusion-related variables and TMD signs or symptoms were found. In other investigations, additional occlusal variables were identified. However, statistical associations do not prove causality. By using Hill's nine criteria of causation, it becomes apparent that the evidence of a causal relationship is weak. Only bruxism, loss of posterior support and unilateral posterior crossbite show some consistency across studies. On the other hand, several reported occlusal features appear to be the consequence of TMDs, not their cause. Above all, however, biological plausibility for an occlusal aetiology is often difficult to establish, because TMDs are much more common among women than men. Symptom improvement after insertion of an oral splint or after occlusal adjustment does not prove an occlusal aetiology either, because the amelioration may be due to the change of the appliance-induced intermaxillary relationship. In addition, symptoms often abate even in the absence of therapy. Although patients with a TMD history might have a specific risk for developing TMD signs, it appears more rewarding to focus on non-occlusal features that are known to have a potential for the predisposition, initiation or perpetuation of TMDs.