A large cohort consisting of all children born to mothers from community provides ‘natural’ selection into different exposures and is a powerful resource for epidemiological research. A large population-based birth cohort with detailed systematic information already recorded, as part of longitudinal medical care, historical and current school data, detailed birth certificate data and all three resources available for every member of the birth cohort, are extremely rare. Our population-based birth cohort consists of all children born between 1976 and 2000 to mothers residing in Olmsted County, Minnesota, at the time of child's birth (N=39 890). In this paper, we provide a comprehensive report of the method describing the identification, the characteristics and longitudinal follow-up of each child (and family members) from the birth cohort, wealth of complementary resources of data and study measures and designs (retrospective, combined retrospective/prospective). In the last decade or so, we obtained scientific and clinically needed answers for incidence rates, potential risk/protective factors, treatment, comorbidities, outcomes, cost/usage and potential biases (that are always assessed and clinically interpreted) of many developmental learning and behavioural disorders (DLBDs) including learning and attention-deficit/hyperactivity disorders, intellectual disability, speech-language impairment and autism spectrum disorder. Many current and future questions related to DLBDs are remaining to be answered. The Olmsted County Birth Cohort (OCBC) is an example of a comprehensive, contemporary epidemiological research model for the development of similar research infrastructures, and its current and future results are important for replication and comparison with other population-based retrospective and prospective birth cohort studies.