3, 4-Didehydroretinol (DR) metabolism was previously followed in vitamin A (VA)-replete lactating sows. This study followed DR appearance and clearance after dosage in serum and milk during 2 lactation cycles in sows (n = 8) fed VA-free feed for 3 gestation-lactation cycles. During lactations 2 and 3, 35 μmol 3, 4-didehydroretinyl acetate was given orally after overnight food deprivation. Blood and milk were collected at 0, 1.5, 3, 5, 7, 9, 16, 24, 36, 48, 60, and 72 h; livers were obtained at kill. Samples were analyzed for DR, retinol (R), and 3, 4-didehydroretinyl esters. During lactations 2 and 3, the 5-h serum DR:R ratios were 0.028 ± 0.017 and 0.069 ± 0.042, respectively, and serum R concentrations were 0.75 ± 0.23 and 0.86 ± 0.37 μmol/L, respectively. The DR:R ratio and serum R were 0.018 ± 0.013 and 0.94 ± 0.12 μmol/L, respectively, in VA-replete sows from the same herd. After lactation 3, liver VA was 0.23 ± 0.05 μmol/g, indicating low-normal VA status. Serum DR area-under-the curve from 0 to 48 h increased as liver stores decreased. Thirteen to 23% of DR dose was secreted into milk, consistent with VA-replete sows. Milk DR concentrations were greater during lactation 3 than 2. Peak concentration occurred earlier and the half-life was shorter for milk DR in the more VA-depleted sows. The milk and serum DR:R were correlated from 3 to 9 h (r = 0.70; P < 0.0001) and increased as VA stores decreased regardless of serum R concentration. Milk DR:R may replace serum measurements during lactation.