In tissues of freshwater fish–feeding mammals, 3,4-didehydroretinol (A2) is a major form of vitamin A. In mink liver, with organochlorine exposure, this analog has been found to decrease more than retinol (A1) and thus has potential as a sensitive freshwater biomarker. The presence of the analogs A1 and A2 as alcohol and different fatty acyl esters, which react to polychlorinated biphenyls differently, necessitates detailed analyses achieved by using direct extraction of tissue homogenate. In direct hexane extraction, compared to total levels of the vitamins obtained in the saponification procedure, a large proportion of the vitamins was released only after repeated and long-time vortex mixing with the extraction solvent. Thus, in tissue extraction, the use of internal standardization alone can lead to a rough underestimation of the levels of these fat-soluble vitamins. For analyses of vitamins A1 and A2 in liver, we applied the argentation high-performance liquid chromatography, which provided good separation of individual A1 and A2 fatty acyl esters. We report retention times for numerous esters of A1 and A2 and, to aid identification, the change in their retention properties after adding AgNO3 to the mobile phase. The argentation did not affect the recoveries of any forms of the retinoids studied but destroyed half the vitamin E. Despite selective acylation of fatty acids into the vitamin A esters, the fatty acids of the esters were the same as those found to be the major fatty acids in the gas–liquid chromatography of total lipids. The goal of this work was to create a methodology that is suitable for biomonitoring alcoholic and esterified vitamins A1 and A2 in tissues of freshwater fish–feeding mammals.