Electrophoretic studies of proteins remain a primary source of insight into genetic diversity in many species including the Atlantic salmon Salmo salar, one of the most culturally and economically important fish species of the North Atlantic region. Since 1966, >350 scientific papers on protein variation have been published encompassing 25 000þ salmon from over 400 locations in >200 river systems across the species' distribution. Variation has been detected at 30% of the 110 protein loci screened, though most studies examine <40. The method has been applied largely to the investigation of population structure and differentiation, but work has also led to the systematic revision of the genus Salmo and remains the primary source of insight into hybridization in the wild with brown trout Salmo trutta. Spatial patterns of differentiation show temporal stability, both within and among river systems, and strongly support structuring of the species into river and tributary specific populations and the designation of European and North American populations as distinct sub-species. They also show widespread regional differentiation within both continents, beyond the marked subcontinental differences between Baltic Sea and Atlantic Ocean populations in Europe. Most of the differentiation probably reflects gene flow and founder events associated with colonization following the retreat of the glaciers from much of the species' modern range. However, variation at MEP-2* shows strong correlations with environmental temperature, both within and among rivers, and associations with phenotypic performance. This suggests selection is acting on the locus and provides †Author to whom correspondence should be addressed. Tel.: þ44 (0) 796 472060; fax: þ44 (0) 1796 473523; # 2005 Crown copyright compelling evidence for the local adaptation of populations. Protein studies have led to more population centred management of the species and have been exploited in the discrimination of regional stocks in mixed stock analysis in high seas fisheries, particularly in the Baltic Sea, and as markers for the assessment of stocking success. They have also advanced insight into how the genetic character of populations can be changed in cultivation and the potential impact of salmon aquaculture and stocking on wild populations. The method has been largely superseded by DNA based analyses, but the results remain highly relevant to Atlantic salmon management and conservation and are an irreplaceable data set for studying genetic stability of populations over time. # 2005 Crown copyright