Associations of serotonin 2A receptor (5-HTR2A) gene polymorphisms with clinical response to atypical antipsychotics (AAPs) treatment in schizophrenia (SCZ) were inconsistent. Thus we conducted a meta-analysis to investigate more reliable estimates. The Cochrane Library, Embase, PubMed, Weipu, CNKI and Wanfang databases were searched for eligible studies published up to September 2021. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) were calculated in four genetic models. Subgroup analyses were performed by ethnicity and antipsychotic type. Meta-regression was used to evaluate the potential effects of confounding variables. In total, 19 studies were included for the meta-analysis, of which 17 studies containing 2359 patients were identified for T102C polymorphism and 7 studies containing 1408 patients for A-1438G polymorphism. The results showed that A-1438G polymorphism was significantly associated with clinical response to AAPs treatment in SCZ in four genetic models (allele model, A vs. G, OR = 1.87, 95% CI = 1.05–3.33, P = 0.034; recessive model, AA vs. GA + GG: OR = 1.79, 95% CI = 1.17–2.72, P = 0.007; dominant model, AA + GA vs. GG: OR = 3.40, 95% CI = 1.15–10.10, P = 0.027; co-dominant model, AA vs. GG: OR = 3.44, 95% CI = 1.07–11.10, P = 0.039) in Asians, but not in Caucasians. When stratified by antipsychotic type, A-1438G polymorphism was related to the efficacy of olanzapine in recessive model (AA vs. GA + GG, OR = 1.85, 95% CI = 1.18–2.90, P = 0.007), but not in other models. However, neither four genetic models nor subgroup analyses of T102C polymorphism were found any significant associations with AAPs response (P > 0.05). Meta-regression revealed that no association was confounded by mean age, male ratio, treatment duration and illness duration (P > 0.05). The present meta-analysis indicated that 5-HTR2A A-1438G polymorphism, but not T102C polymorphism, was significantly associated with AAPs response in SCZ, especially in Asians and olanzapine-treated patients.