The use of assisted reproductive technologies (ART) for the treatment of infertility has grown exponentially over the last 20 years, and now accounts for 4% of all births in Australia, and over 1 m births annually around the globe. There is consistent reporting of increased risk of adverse perinatal outcomes and birth defects following infertility treatment. However, change in practice has been stymied by critical knowledge gaps with regards to (a) the relative contribution of patient and treatment factors to adverse outcomes, (b) the independent contribution of specific contemporary treatments and treatment combinations to outcomes, (c) the impact of innovations in laboratory and clinical practice on treatment success and observed risk and (d) changes over time in patient characteristics. Here we summarize key findings from the South Australian Birth Cohort, which is a whole-of-population cohort of over 300,000 births from 1986 to 2002. Relative to spontaneous conceptions, singletons from assisted conception were more likely to be stillborn [odds ratio (OR)=1.82; 95% confidence interval (CI) 1.34–2.48], while survivors as a group were comprehensively disadvantaged at birth, including lower birth weight (OR=2109 g; 95% CI 2129–289), very low birth weight (OR=2.74; 95% CI 2.19–3.43), very preterm birth (OR=2.30; 95% CI 1.82–2.90) and neonatal death (OR=2.04; 95% CI 1.27–3.26). Major birth defects, including cardiac, urogenital and musculoskeletal defects are doubled after fresh ICSI cycles, which is a particular concern as ICSI now accounts for 70% of all treatment cycles globally. Future study is needed to provide contemporary, precise evidence to inform patient and clinic decision making, and generate knowledge for future innovation in ART laboratory methods and clinical practice, thereby optimizing treatment and health outcomes while reducing adverse events.