This meta-analysis aims to evaluate the efficacy and safety of antiprogressive disease (PD)-(L)1-based neoadjuvant therapy in head and neck squamous cell carcinoma (HNSCC) patients and identify potential prognostic biomarkers. Databases were systematically searched for prospective clinical trials evaluating the efficacy and safety of anti-PD-(L)1-based neoadjuvant therapy for HNSCC before January 12, 2024. We estimated the efficacy and safety of neoadjuvant immune checkpoint inhibitors. Subgroup and sensitivity analyses were further performed. A total of 570 patients from 20 studies were included. The pooled major pathological response (MPR), pathological complete response (pCR), and partial pathological response (PPR) rates were 30.7%, 15.3%, and 68.2%, respectively. Surgical complications, surgical delayed rate, all grade treatment-related adverse effects (TRAEs) and ≥Grade 3 TRAEs were 0.6%, 0.3%, 82.6%, and 9.7%, respectively. Best MPR or pCR rate was detected in patients receiving neoadjuvant anti-PD-(L)1 therapy + radiotherapy (with MPR rate of 75.5% and pCR rate of 51.1%) and neoadjuvant anti-PD-(L)1 therapy + chemotherapy groups (with MPR rate of 57.5% and pCR rate of 26.7%). No differences were detected in subgroups stratified by neoadjuvant treatment cycles, human papillomavirus (HPV) status, and tumor location. Patients with baseline Combined Positive Score (CPS) ≥ 20 have higher MPR and pCR rates compared to patients with CPS < 20. High Tumor Cell Proportion Score was also associated with MPR and pCR. Objective response rate is a strong predictor of MPR (odds ratio [OR] = 7.78, 95% confidence interval [CI] = 3.20%-18.91%) and pCR (OR = 3.24, 95% CI = 1.40%-7.48%). Anti-PD-(L)1-based neoadjuvant therapy was effective and safe for HNSCC patients.