Aim To determine the age and ethnicity-specific prevalence of anisometropia in Australian preschool-aged children and to assess in this population-based study the risk of anisometropia with increasing ametropia levels and risk of amblyopia with increasing anisometropia.

Methods A total 2090 children (aged 6–72 months) completed detailed eye examinations in the Sydney Paediatric Eye Disease Study, including cycloplegic refraction, and were included. Refraction was measured using a Canon RK-F1 autorefractor, streak retinoscopy and/or the Retinomax K-Plus 2 autorefractor. Anisometropia was defined by the spherical equivalent (SE) difference, and plus cylinder difference for any cylindrical axis between eyes.

Results The overall prevalence of SE and cylindrical anisometropia ≥1.0 D were 2.7% and 3.0%, for the overall sample and in children of European-Caucasian ethnicity, 3.2%, 1.9%; East-Asian 1.7%, 5.2%; South-Asian 2.5%, 3.6%; Middle-Eastern ethnicities 2.2%, 3.3%, respectively. Anisometropia prevalence was lower or similar to that in the Baltimore Pediatric Eye Disease Study, Multi-Ethnic Pediatric Eye Disease Study and the Strabismus, Amblyopia and Refractive error in Singapore study. Risk (OR) of anisometropic amblyopia with ≥1.0 D of SE and cylindrical anisometropia was 12.4 (CI 4.0 to 38.4) and 6.5 (CI 2.3 to 18.7), respectively. We found an increasing risk of anisometropia with higher myopia ≥−1.0 D, OR 61.6 (CI 21.3 to 308), hyperopia > +2.0 D, OR 13.6 (CI 2.9 to 63.6) and astigmatism ≥1.5 D, OR 30.0 (CI 14.5 to 58.1).

Conclusions In this preschool-age population-based sample, anisometropia was uncommon with inter-ethnic differences in cylindrical anisometropia prevalence. We also quantified the rising risk of amblyopia with increasing SE and cylindrical anisometropia, and present the specific levels of refractive error and associated increasing risk of anisometropia.