The objective was to explore how perinatal mortality relates to birthweight, gestational age and optimal perinatal survival weight for two Arctic populations employing an existing and a newly established birth registry. A medical birth registry for all births in Murmansk County of North-West Russia became operational on 1st January 2006. Its primary function is to provide useful information for health care officials pertinent to improving perinatal care. The cohort studied consisted of 17,302 births in 2006-07 (Murmansk County) and 16,006 in 2004-06 (Northern Norway). Birthweight probability density functions were analysed, and logistic regression models were employed to calculate gestational-age-specific mortality ratios. The perinatal mortality rate was 10.7/1000 in Murmansk County and 5.7/1000 in Northern Norway. Murmansk County had a higher proportion of preterm deliveries (8.7%) compared to Northern Norway (6.6%). The odds ratio (OR) of risk of mortality (Northern Norway as the reference group) was higher for all gestational ages in Murmansk County, but the largest risk difference occurred among term deliveries (OR 2.45, 95% confidence interval 1.45, 4.14) which hardly changed on adjustment for maternal age, parity and gestation. Proportionately, more babies were born near (± 500 g) the optimal perinatal survival weight in Murmansk County (67.2%) than in Northern Norway (47.6%). The observed perinatal mortality was higher in Murmansk County at all birthweight strata and at gestational ages between weeks 25 and 42, but the adjusted risk difference was most significant for term deliveries.