Objectives Prenatal exposure of long chain polyunsaturated fatty acids (LCPUFA) are essential for normal fetal growth and neurodevelopment. Availability of LCPUFA depends mostly on maternal fish consumption. Fish consumption also exposes the fetus to mercury which is well known neurotoxicant. We analyzed the associations of combined LCPUFA and mercury from fish consumption during pregnancy on newborn's brain measures and child neurodevelopment in a northern Adriatic coastal area. Patients and methods The prospective cohort study included 257 mother – infant pairs enrolled in a susceptible population of the Public Health Impact on long- term, low-level, Mixed Element exposure (PHIME) EU Sixth Framework Programme from 2 recruitment areas of the northern part of the Adriatic coast. Umbilical cord blood taken at delivery was used for measuring concentration of total mercury (THg) and specific LCPUFA - docosahexaenoic acid (DHA) and arachidonic acid (ARA). Neonatal cranial sonography was performed at the age of 3 days in 57 newborns. Neurodevelopmental assessment of cognitive, motor and language skills were conducted at 257 children at the age of 18 months using the Bayley Scales of Infant and Toddler Development, Third Edition. The participants were divided into two groups depending on the THg concentration in the umbilical cord blood (exposed > 5.8 μg/L and unexposed < 5.8 μg/L). Dietary habits and exposures to environmental and social factors were assessed through questionnaires. Results There is a statistically significant difference in the cerebellum length (p = 0.032) and the superior frontal gyrus width (p = 0.023) between the exposed and the unexposed group. In combined analysis, including possible protective variables as DHA and ARA (R2 = 0.22, p = 0.001), the negative contribution of THg on cerebellum length (beta = −0.16, p = 0.001) persisted. We found no correlation between THg concentration in umbilical cord blood and child neurodevelopment scores at the age of 18 months. Language score with receptive and expressive subscores was significantly associated with fish consumption (p < 0.05). Conclusion This analysis demonstrates the existence of morphological brain changes in newborns that are prenatally exposed at mercury concentrations what was diminished in combined analyse including LCPUFA. Our results emphasizes the importance of LCPUFA's and mercury common influence as a predictor of developmental outcome. Fish consumption, not solely LCPUFA contributes to better language development of children at the age of 18 months.