In last few years Italy has faced ad increase in Emergency Department (ED) admissions, mostly for non-urgent conditions: although institutions often attribute this trend to perceived General Practice's (GP) inefficiency, evidence remains scarce.This study analyses inappropriate ED visits, quantifying the proportion of patients bypassing GP consultation and evaluating the GP’s effectiveness as a gatekeeper for emergency services' access.This descriptive observational study examined 300 ED admissions among patients of four GPs in Piacenza, Italy. For each case, we recorded: mode of referral (self-referral vs. GP-mediated or specialist-mediated access), prior contact with GP, triage code and clinical inappropriateness (feasibility of management within Primary Care). Data were analysed using absolute frequencies and proportions.Most patients (84%, n=251) did not consult their GPs before accessing to the ED and 87% (n=261) were self-referred, occasionally even after prior medical consultation with their GPs. Among the 234 self-referred visits without GP contact, only 43% (n=100) were deemed clinically appropriate, compared to 79% (n=34) of the 43 visits filtered by the attending physician. While minor triage codes accounted for 67% (n=200) of cases, clinical judgment determined that only 51% (n=154) of total visits were inappropriate. Notably, only 2% (n=5) of self-presentations followed a lack of response from the GP; among these, 3 cases were assigned "major" triage codes suggesting time-dependent conditions.Approximately half of ED accesses are clinically unnecessary and manageable within a Primary Care setting. High ED utilization is primarily driven by patient self-initiative rather than GP unresponsiveness, which accounted for a negligible fraction of self-refferals. Conversely, when General Practice performs a gatekeeping role, ED utilization is more frequently appropriate. Minor triage codes not always correspond to inappropriateness of access according to clinical judgment (it is standard practice to assign minor codes to obstetric or orthopedic conditions requiring urgent instrumental diagnostics).The patient’s tendency to bypass the GP’s gatekeeping function, rather than a presumed inefficiency of General Practice is the primary driver of ED overcrowding from non-acute presentations. Lack of GP responsiveness is a misconception. Study limitations include: small sample size and potential confirmation bias.