By means of a qualitative study, we set out to generate hypotheses about the way general practitioners (GPs) diagnosed dementia. We looked for triggers for the diagnosis of dementia. Ten GPs were interviewed about four dementia cases taken from their own practices. GPs are more concerned with treatment than with the diagnosis of dementia as such, as diagnosis is performed in relation to treatment. The most important aid was the evolution of the clinical tables based on the information of caregivers. Important triggers were changes of activities of daily life function, behaviour and cognition. Acute illness and loss of the key caregiver were relevatory moments. Recommendations are made for the improvement of the diagnostic abilities of GPs.