Data is presented, which suggest that the day length a woman experiences during the periconceptional period predicts the C677T‐MTHFR genotype of her child. Logistic regression analysis involving 375 neonates born in the same geographical location within a three year period demonstrated that photoperiod (minutes) at conception predicts both genotype (P = 0.0139) and mutant allele carriage (P = 0.0161); the trend clearly showing that the 677T‐MTHFR allele frequency increases as photoperiod increases. We propose a number of explanations, including a hypothesis in which a long photoperiod around conception decreases maternal systemic folate because of UVA induced dermal oxidative degradation of 5‐methyl‐H4folate, leading to a lower cellular 5,10‐methylene‐H4folate status. In this scenario, 5,10‐methylene‐H4folate would be more efficiently used for dTMP and DNA synthesis by 677T‐MTHFR embryos than wildtype embryos giving the 677T‐MTHFR embryos increased viability, and hence increasing mutant T‐allele frequency. Alternate hypotheses include: increased seasonal availability of folate rich foods that genetically buffer any negative effect of 677T‐MTHFR in embryos; seasonal oxidative stress lowering embryo‐toxic homocysteine; an undefined hormonal effect of photoperiod on the neuroendocrine axis, which mediates genotype/embryo selection. The effect of photoperiod on genotype seems clear, but the speculative molecular mechanism underpinning the effect needs careful examination. Am. J. Hum. Biol., 2010. © 2010 Wiley‐Liss, Inc.