To investigate the association between exposure to anesthesia during three periods of pregnancy, delivery, and childhood and autism spectrum disorder (ASD). PubMed, Scopus, Web of Science, Embase, Google Scholar, PsycArticles, and PsycINFO were searched from the date of database inception to 1 December 2022. Studies reported the association between exposure to anesthesia during pregnancy, delivery, and childhood and ASD were included. Extracted variables included hazard ratio (HR), relative risk or odds ratio, standard error, and 95% confidence interval (CI). Effect estimates were pooled using random-effects meta-analysis. In total, 16 studies including 8,156,608 individuals were included in the meta-analysis. Labor epidural anesthesia during delivery was associated with ASD in the general population (adjusted HR = 1.16, 95% CI, 1.06-1.28) but not in the sibling population (adjusted HR = 1.06, 95% CI, 0.98-1.15). Other anesthesia during delivery was not associated with ASD (general population: adjusted HR = 1.08, 95% CI, 0.99-1.17; sibling population: adjusted HR = 1.20, 95% CI, 0.81-1.79). Three studies suggested that exposure to anesthesia during pregnancy was associated with ASD in offspring (adjusted HR = 2.15, 95% CI, 1.32-3.48). There was no significant association between exposure to general anesthesia during childhood and ASD (adjusted HR = 1.02, 95% CI, 0.60-1.72). This meta-analysis did not confirm the association between exposure to anesthesia during labour and ASD. Previous observational studies used the neurotoxicity of anesthesia to biologically explain significant associations, but in fact different controls for confounding factors led to differences in associations. The evidence for pregnancy and childhood was limited given the small number of studies in these periods.