No study has investigated the association between number of anxiety disorders (NADs) and long-term outcome over 10 years among patients with major depressive disorder (MDD). This study investigated this issue.At baseline, 290 outpatients with MDD were enrolled, 149 with at least one anxiety disorder (AD). Subjects were followed-up at six-month, two-year, and 10-year points. The Structured Clinical Interview for DSM-IV-TR was used to confirm psychiatric diagnoses. NADs at baseline was recorded. The Hamilton Depression Rating Scale (HAMD), the anxiety subscale of the Hospital Anxiety and Depression Scale (HADS-A), and the somatic subscale (SS) of the Depression and Somatic Symptoms Scale were employed. Generalized Estimating Equation models were used for statistical analysis.MDD patients with ADs had greater depression, anxiety, and somatic severities at the three follow-up points than those without. NADs was significantly and positively correlated with the three dimensions and total duration of pharmacotherapy at follow-up. NADs was independently associated with symptom severity after controlling for depression and anxiety at baseline and pharmacotherapy. When the DSM-5 criteria for ADs were applied, the results were unchanged. Specific phobia, panic disorder and social phobia, and panic disorder and specific phobia were independently associated with depression, anxiety, and somatic symptoms, respectively.Pharmacotherapy at follow-up was not controlled. The three follow-up intervals were unequal.Comorbidity with more ADs was associated with a poorer prognosis. The negative impacts of ADs on MDD persisted at the ten-year follow-up point. NADs was associated with the long-term prognosis of MDD.