Abstract Purpose To investigate omission of population‐based eye screening at age 6–9 months in the Netherlands. Methods Prospective population‐based consecutive birth cohort study was used. In two consecutive birth cohorts, children were eye screened at 1–2 and 3–4 months, but at general‐health screening at 6–9 months, the second cohort was not eye screened, unless anything conspicuous was noted or in case of positive family history. Data were collected from screening records and anonymous questionnaires. Semi‐structured daylong observations were made of physicians examining children aged 0–4 years, including children from the cohorts, by two orthoptic students. Results 58 of 6059 children (0.96%), in the screened, and 48 of 5482 children (0.88%) in the unscreened group were referred to orthoptist or ophthalmologist, mostly for observed strabismus. Amblyopia, all combined with strabismus, was diagnosed in ten screened (0.17%) versus six unscreened children (0.11%). Most physicians found preverbal examinations and decisions to refer difficult. The observations by orthoptic students revealed that cover test, pupillary reflexes, pursuit movements and eye motility were frequently performed inadequately, contrary to the Hirschberg test, at this age. Conclusion The screened and unscreened group differed little regarding the number of children referred and found to have amblyopia. Referral was mostly based on observed strabismus.