The chief aim of this research is to investigate the causality of air pollutants and adverse pregnancy outcomes. Two-sample Mendelian randomization was conducted, employing genetic variants connected with air pollution as instrumental variables. Sixteen adverse pregnancy outcomes were extracted as the main outcome measures from the genome-wide association study (GWAS). The inverse-variance weighted (IVW) method was conducted as the primary analysis method. This study found that there were causal association between NO2 and pre-eclampsia (weighted median: OR = 1.30, 95% CI = [1.03–1.64], p = 0.029) and between PM2.5 and placental abruption (IVW: OR = 10.94, 95% CI = [1.28–93.45], p = 0.029). There were potential causal relationships between NO2 and gestational hypertension (IVW: OR = 1.14, 95% CI = [0.99–1.30], p = 0.060); NO2 and placental abruption (IVW: OR = 1.97, 95% CI = [0.90–4.28], p = 0.089); NOx and fetal growth restriction (IVW: OR = 0.06, 95% CI = [0.99–1.12], p = 0.089); PM2.5 and slow fetal growth and fetal malnutrition (MR–Egger: OR = 54,240.95, 95% CI = [2.08–1,411,757,729.46], p = 0.059); PM10 and hyperemesis gravidarum (MR–Egger: OR = 0.12, 95% CI = [0.02–0.97], p = 0.086); PM10 and preterm birth (weighted median: OR = 1.60, 95% CI = [0.95–2.70], p = 0.075); and PM10 and spontaneous abortion (weighted median: OR = 1.60, 95% CI = [0.95–2.70], p = 0.075). There was no pleiotropy, but there was some heterogeneity. In conclusion, air pollution has a causal effect on several adverse pregnancy outcomes.