<p>The adverse neurological effects of methylmercury are well characterized in animal models and humans. Neurodevelopmental effects in infants and children following prenatal exposure to methylmercury have been the critical endpoint used for many years in risk assessment for the derivation of toxicological reference values. Most risk assessments of methylmercury have continued to rely on the same epidemiological birth-cohorts from New Zealand, the Faroe Islands, and the Seychelles, in which neurodevelopmental effects of methylmercury were examined during the 1980s and 1990s. New epidemiological studies have since emerged, prompting several organizations to reassess whether current reference values remain protective. As these evaluations proceed, clarity on the basis for derivation of the existing values will be critical particularly if different toxicological reference values are recommended. This review summarizes the history, pivotal data, and methodologies used by different regulatory bodies and expert committees (US EPA, ATSDR, Health Canada, JECFA, and EFSA) to derive toxicological reference values for neurodevelopmental effects of prenatal methylmercury exposure. Although methodologies differed in the choice of points of departure, uncertainty factors, and dosimetry conversions, there is good agreement between the toxicological reference values derived by these organizations. Nonetheless, there remain important uncertainties and data gaps particularly related to how observational epidemiology data were used in derivation of the toxicological reference values. Discussion of these uncertainties should help inform future work on methylmercury and have broader implications for other chemical risk assessments relying on similar types of observational data.</p>