Fish is an important causative material of food allergy. Although the allergenicity of fish is considered to correlate with the content of parvalbumin, the major fish allergen, available information about the parvalbumin content in fish is limited. In this study, a simple and reliable quantification method for fish parvalbumin by SDS–PAGE was first established. Application of the SDS–PAGE method to 22 species of fish revealed a marked variation in parvalbumin content among fish. Furthermore, the parvalbumin content was found to be higher in dorsal white muscle than in ventral white muscle, in rostral part of white muscle than in caudal part of white muscle and in white muscle than in dark muscle. IgE reactivity of fish was roughly proportional to parvalbumin content. Interestingly, large-sized migratory fish, such as salmon, swordfish and tuna, were commonly very low in both parvalbumin content and IgE reactivity.