The aim of this study was to investigate the association of breastfeeding duration and food variety at 2 years of age. A secondary data analysis was undertaken of the Western Australian Pregnancy Cohort (Raine) Study, an ongoing longitudinal study. Data collected from a single 24 h dietary recall of 1905, 2 year-old children were used to calculate two food variety scores; a core food variety score (CFVS) and a fruit and vegetable variety score (FVVS). Multivariate linear regression analysis was used to identify those factors independently associated with the CFVS and FVVS. The mean CFVS was 7.52 (range 1–18) of a possible 34 food items or groups and the mean FVVS was 2.84 (range 0–10) of a possible 16 food items or groups. Breastfeeding duration was independently directly associated with the CFVS (p < 0.001) and FVVS (p < 0.001). In addition, maternal age was independently directly associated with the CFVS (p < 0.001) and FVVS (p = 0.001) as was maternal education (CFVS p < 0.001 and FVVS p = 0.043). The presence of older siblings was independently inversely associated with the CFVS (p = 0.003) and FVVS (p = 0.001). This study demonstrated a direct modest association between breastfeeding duration and food variety in 2 year-old children, independent of maternal demographic characteristics known to predict food variety in children. This finding supports the hypothesis that flavours transferred in breast milk provide repeated early exposure to different tastes and positively shape children’s food preferences and food variety.