Lactation hampers normalization of the maternal arachidonic acid (AA) status, which is reduced after pregnancy and can further decline by the presently recommended increased consumption of (n-3) long-chain PUFA [(n-3) LCPUFA]. This may be unfavorable for breast-fed infants, because they also require an optimum supply of (n-6) LCPUFA. We therefore investigated the LCPUFA responses in nursing mothers upon increased consumption of AA and (n-3) LCPUFA. In a parallel, double-blind, controlled trial, lactating women received for 8 wk no extra LCPUFA (control group, n = 8), 200 (low AA group, n = 9), or 400 (high AA group, n = 8) mg/d AA in combination with (n-3) LCPUFA [320 mg/d docosahexaenoic acid (DHA), 80 mg/d eicosapentaenoic acid, and 80 mg/d other (n-3) fatty acids], or this dose of (n-3) LCPUFA alone [DHA + eicosapentaenoic acid group, n = 8]. Relative concentrations of AA, DHA, and sums of (n-6) and (n-3) LCPUFA were measured in milk total lipids (TL) and erythrocyte phospholipids (PL) after 2 and 8 wk and changes were compared by ANCOVA. The combined consumption of AA and (n-3) LCPUFA caused dose-dependent elevations of AA and total (n-6) LCPUFA concentrations in milk TL and did not significantly affect the DHA and total (n-3) LCPUFA increases caused by (n-3) LCPUFA supplementation only. This latter treatment did not significantly affect breast milk AA and total (n-6) LCPUFA concentrations. AA and DHA concentrations in milk TL and their changes were strongly and positively correlated with their corresponding values in erythrocyte PL (r(2) = 0.27-0.50; P </= 0.002). We thus concluded that the consumption by lactating women of AA in addition to extra (n-3) LCPUFA dose dependently increased the AA concentration of their milk TL.