Abstract Background Most patients with esophageal cancer (EC) present locally advanced disease. Currently, multimodal strategies as neoadjuvant chemoradiotherapy or perioperative chemotherapy are employed, but still have high local and distant failures rates. This study evaluated the results of a phase II protocol with induction chemotherapy (IC) and concurrent chemoradiation (CR) before surgery for locally advanced carcinoma of the esophagus and esophagogastric junction (EGJ). The primary end point was to evaluate pathological complete remission (PCR) rate at surgical specimen. A PCR ≥ 30% was considered promising for further study. Methods Patients with squamous cell carcinoma (SCC) or adenocarcinoma (AC) of the thoracic esophagus and JEG, clinical stage T1b-3, N0-2, M0, PS 0-2, underwent two cycles of IC with carboplatin and paclitaxel followed by CR with the same agents and minimally invasive esophagectomy. At the screening, 14 days after the first cycle of IC and 6-8 weeks after the CR, metabolic responses were accessed by PET-CT and classified according the PERCIST criteria. Results One hundred and twenty-nine patients were accrued from January 2017 to April 2023. Forty-seven patients started the trial and 45 completed the neoadjuvant part of the protocol. Most patients were male (73.3%), the median age of was 55 years and AC was the most prevalent histological subtype (53.3%). Grade 3/4 toxicity during IC were infrequent and included neutropenia (12.5%), anemia (6.2%) and dysphagia (10.4%). Lymphopenia (10.7%) and dysphagia (6.5%) were the most common grade 3/4 toxicity during CR. All cases received the full radiotherapy dose of 45 Gy by IMRT technique. Thirty-seven patients underwent surgery and 31 had minimally invasive MacKeown esophagectomy with complete tumor resection (R0). Sixty-four percent of all included patients completed the neoadjuvant therapy and survived for 30 days after surgery. The most prevalent postoperative complications were anastomotic fistula (29%) and pneumonia (19%). Median hospital stay was 12 (7-56) days. Six patients (19.3%) had major postoperative morbidity (Clavien-Dindo ≥ IIIb), 4 were reoperated and one (3.2%) died during hospitalization. The median number of lymph nodes retrieved was 24. Eleven patients (33.5%) had pathological positive lymph nodes. The overall pathological complete response rate was 25.8%, 41.8% for SCC and 15.8% for AC. The early metabolic evaluation by PERCIST showed that 2 cases (4.4%) had progressive disease (PD), 26 (57.7%) stable disease (SD), 15 (33.3%) partial response (PR) and 2 (4.4%) complete response (CR). All the CR were SCC. The late metabolic response analysis showed that 1 (2.4%) patient had PD, 6 (14.6%) SD, 23 (56%) PR and 11 (27%) CR. On the basis of an intention-to-treat analysis, the median overall survival was 31.4 months, whereas the five-year overall survival was 31.2%. After curative surgery, 12 patients had recurrence disease. Six (19%) patients had recurrences at distant sites, 4 (12.9%) had loco-regional recurrence and 2 (6.4%) had both patterns. Conclusion This intensified neoadjuvant regimen was efficacious and safe in patients with locally advanced EC, but failed to reach the defined efficacy goal for further investigation, especially for AC.