It is established that delayed effective pharmacotherapy plays a significant role in the overall burden of psychiatric disorders, which are often treated with symptomatic drugs, that is benzodiazepines (BZDs), in relation to their rapid onset of action and safety, despite long-term side effects. We aimed to assess the influence of initial treatment with BZDs on the duration of untreated illness (DUI) and whether specific sociodemographic and clinical factors could influence the choice of BZDs as first treatment in 545 patients affected by schizophrenia, mood and anxiety spectrum disorders. Statistical analyses (one-way analysis of variance and χ2) were carried out to compare patients who used BZDs as first treatment (BZD w/) and those who did not (BZD w/o). The overall DUI, irrespective of diagnosis, resulted in significantly longer in BZD w/ versus w/o patients, who also experienced more frequently anxious/depressive symptoms at onset. Furthermore, BZD w/ patients more frequently autonomously decided to look for treatment (mainly refering to psychologists or general practitioners) and experimented more frequently phobias, than BZD w/o ones. The present findings suggest that initial BZDs treatment may prolong the overall DUI, although their prescription seems to be influenced by specific sociodemographic and clinical factors. Further studies are needed to confirm the present findings.