Prevalence and risk factors associated with mixed anxiety-depressive disorder (MAD) have yet to be established. Using MINI 5.0.1 and HADS, a two-week survey involving 21,644 primary care patients was carried out. We found 1.8% of subjects with MAD and 20% of subjectswith a co-morbid anxiety and depression (CAD) disorder. MAD patients without a past history of anxiety/affective episodes were defined as "pure MAD" (pMAD: 0.9% of the sample). While MAD patients showed a number of differences vs. the other groups of patients in the socio-demographic statistics, pMAD patients were not different, apart from a higher proportion of males vs. CAD patients. Nearly in all the comparisons, MAD and pMAD patients showed lower association with life events and with a familial predisposition than the other patients. On HADS assessment, MAD showed a higher risk of anxiety and depressive symptoms than anxiety diagnoses, a lower risk of depressive symptoms than depressive diagnoses and a lower risk of both anxiety and depressive symptoms than CAD. Since more than a half of MAD patients were classified as pMAD, the hypothesis that MAD should be viewed as a partial remission of a major depression is not entirely confirmed in our study.