Basal information of head and neck small-cell carcinoma (HNSmCC) including epidemiology, primary site, treatment, and prognosis remains sparse due to its rarity. We report here a multicenter retrospective study on the diagnosis, treatment, and outcomes of patients with HNSmCC. This study involved 47 patients with HNSmCC from 10 participating institutions. Eight patients were excluded for whom no pathological specimens were available (n = 2) and for discrepant central pathological judgements (n = 6). The remaining 39 patients were processed for data analysis. As pretreatment examinations, computed tomography (CT) was performed for the brain (n = 8), neck (n = 39), and chest (n = 32), magnetic resonance imaging (MRI) for the brain (n = 4) and neck (n = 23), positron emission tomography-CT (PET-CT) in 23 patients, bone scintigraphy in 4, neck ultrasonography in 9, and tumor markers in 25. Primary sites were oral cavity (n = 1), nasal cavity/paranasal sinuses (n = 16), nasopharynx (n = 2), oropharynx (n = 4), hypopharynx (n = 2), larynx (n = 6), salivary gland (n = 3), thyroid (n = 2), and others (n = 3). Stages were II/III/IV-A/IV-B/IV-C/Not determined = 3/5/16/6/5/4; stage IV comprised 69%. No patient had brain metastases. First-line treatments were divided into 3 groups: the chemoradiotherapy (CRT) group (n = 27), non-CRT group (n = 8), and best supportive care group (n = 4). The CRT group included concurrent CRT (CCRT) (n = 17), chemotherapy (Chemo) followed by radiotherapy (RT) (n = 5), and surgery (Surg) followed by CCRT (n = 5). The non-CRT group included Surg followed by RT (n = 2), Surg followed by Chemo (n = 1), RT alone (n = 2), and Chemo alone (n = 3). The 1-year/2-year overall survival (OS) of all 39 patients was 65.3/53.3%. The 1-year OS of the CRT group (77.6%) was significantly better compared with the non-CRT group (31.3%). There were no significant differences in adverse events between the CCRT group (n = 22) and the Chemo without concurrent RT group (n = 9). Neck and chest CT, neck MRI, and PET-CT would be necessary and sufficient examinations in the diagnostic set up for HNSmCC. CCRT may be recommended as the first-line treatment. The 1-year/2-year OS was 65.3%/53.3%. This study would provide basal data for a proposing the diagnostic and treatment algorithms for HNSmCC.