This study aimed to evaluate the association between parental and infant birth weights in Japan. In total, 37,504 pregnant Japanese women and their partners were included in this birth cohort study. A multinomial logistic regression model was used to evaluate the associations of parental birth weights with small-for-gestational-age (SGA) or large-for-gestational-age (LGA) infants. Associations between parental birth weight and low birth weight (LBW) infants or macrosomia were also examined, and linear associations between parental birth weight and SGA or LGA were found. The adjusted odds ratios (aORs) for SGA infants per 500 g decrease in maternal and paternal birth weights were 1.50 (95% confidence interval [CI],1.43–1.58) and 1.31 (95% CI, 1.25–1.38), respectively. The aORs for LGA infants per 500 g increase in maternal and paternal birth weights were 1.53 (95% CI, 1.47–1.60) and 1.41 (95% CI, 1.35–1.47), respectively. The association between parental birth weight and LBW infants or macrosomia was also linear. The aORs for LBW infants per 500 g decrease in maternal and paternal birth weights were 1.47 (95% CI, 1.40–1.55) and 1.25 (95% CI, 1.19–1.31), respectively. The aORs for macrosomia per 500 g increase in maternal and paternal birth weights were 1.59 (95% CI, 1.41–1.79) and 1.40 (95% CI, 1.23–1.60), respectively. Parental birth weight was found to be associated with infant birth weight even after adjusting for various parental factors. Furthermore, maternal birth weight was more strongly associated with infant birth weight than with paternal birth weight.