Background: Prenatal anaemia causes serious consequences for both mother and foetus, and dietary factors are suggested to be associated with anaemia. However, research in pregnant women living in rural areas is limited. We aim to assess the contribution of dietary diversity to the magnitude of prenatal anaemia in rural China and identify the interactions between dietary diversity and several sociodemographic and maternal characteristics in relation to anaemia. Methods: A multi-stage random cluster sampling method was used to select pregnant women in rural western China. The Woman’s Dietary Diversity Score was created to measure dietary diversity, which was recoded into terciles. Multinomial logistic regression models were used to assess the associations between dietary diversity score terciles and the magnitude of prenatal anaemia. Multiplicative interactions were tested by adding the product term of dietary diversity and several sociodemographic and maternal characteristics into the regression models. Results: Out of 969 participants, 54.3% were measured as anaemic, with 28.6% mildly anaemic and 25.7% moderately to severely anaemic. There was an absence of agreement between self-reported and measured anaemia status (κ = 0.28, 95% CI [0.22–0.34]). Participants in the highest dietary diversity score tercile had lower odds of being moderately to severely anaemic after adjusting for potential confounders (RRR = 0.65, 95% CI [0.44, 0.98]). In participants with moderate to severe anaemia, significant interactions were found between dietary diversity score terciles, age, and parity (p for interaction < 0.05). Conclusions: The prevalence of prenatal anaemia in rural China remains high, and pregnant women living in these areas are insufficiently aware of their anaemia status. Improving dietary diversity is needed to manage prenatal anaemia in rural areas.