Two experiments were conducted with the aim of studying how dietary fat source, reproductive stage (Exp. 1), and diurnal variation (Exp. 2) affect plasma ketone bodies in sows. In Exp. 1, 40 second-parity sows were fed 1 of 5 lactation diets from 7 d prepartum until 28 d postpartum, with low or high levels (3% or 8%) of dietary fats with different proportions of medium- and long-chain fatty acids. Blood was obtained by jugular venipuncture on d 3 and 7 prepartum, and d 1, 10, 17, and 28 postpartum, and concentrations of plasma beta-hydroxy butyric acid (BHBA), acetoacetate + acetone (AcAc+Ac), glucose, NEFA, lactate, acetate, and butyrate were determined. For 4 out of 5 treatments, plasma BHBA decreased slightly, whereas plasma AcAc+Ac remained stable. However, plasma BHBA (P < 0.01) and AcAc+Ac (P < 0.001) doubled after d 10 of lactation in sows fed 4% octanoic acid and 4% fish oil diet (4+4% FO; P < 0.001), compared with earlier in lactation (P < 0.001). Plasma AcAc+Ac was positively related to BHBA (P < 0.01), glucose (P < 0.05), and butyrate (P < 0.001), and negatively related to the acetate:butyrate ratio (P < 0.001). In addition, plasma BHBA was positively related to lactate (P < 0.01), acetate, and butyrate (P < 0.05). In Exp. 2, diurnal variations of plasma metabolites were studied in 5 sows sampled every second hour from a jugular catheter throughout a 24-h period on d 5 and 17 of lactation and analyzed as in Exp. 1. In addition, milk and urine samples were collected and analyzed for BHBA and AcAc+Ac. No diurnal variations in plasma BHBA or AcAc+Ac were observed and plasma AcAc+Ac was unchanged from d 5 to 17 of lactation (3.7 μM), whereas BHBA declined from 58 μM on d 5 of lactation to 52 μM on d 17 of lactation (P < 0.05). Minor amounts of AcAc+Ac were found in urine (8.6 μM) and this was not affected by days in milk, whereas the content of AcAc+Ac in milk and BHBA in milk and urine were less than the detection limit in 4 of 5 sows. In conclusion, dietary fat source affected plasma concentrations of ketone bodies, but the concentrations were much less than normally observed in dairy cows and, therefore, primary ketosis does not appear to be a major problem in sows. In addition, this study indicates that the intermediary metabolism of sows was challenged when sows were exposed to high fat diets in late gestation.