We compared the effects on plasma lipids of margarines containing either a trans fatty acid- (TFA)-free hard fraction achieved through interesterification (from primarily saturated fatty acids) or a partially hydrogenated hard fraction. Thirty-eight mildly hyperlipidemic subjects consumed a low-fat diet for 2 wk. They were then allocated to 2 groups and underwent 3 dietary interventions for 3 wk each in random order (diets contained 35% of energy as fat with 20% of energy as margarines or butter): group 1 (n = 18), butter, canola oil blend with TFAs, and TFA-free canola oil blend: group 2 (n = 19), butter, polyunsaturated oil blend with TFAs, and TFA-free polyunsaturated oil blend. Plasma LDL-cholesterol concentrations after consumption of all oil blends and after the low-fat diet were not significantly different, but were 11-15% lower than after butter (P < 0.001). Whereas the canola oil blends did not differ from each other in their effects on lipoprotein profiles, the TFA-free polyunsaturated oil blend resulted in a significant 6% reduction (-0.25 mmol/L; 95% CI: -0.08, -0.42) in total and LDL cholesterol compared with the blend containing TFAs (P = 0.006). In the canola oil blend, 10% TFAs and 6% oleic acid were replaced by approximately 14% saturated fatty acids plus a 2% increase in linoleic acid. In the polyunsaturated oil blend, 10% TFAs and 5% linoleic acid were replaced by a 10% increase in saturated fatty acids as well as 3% and 1% increases in oleic and linolenic acids, respectively. We conclude that, compared with butter, TFA-free margarines may be equal to or more effective than margarines containing TFAs in lowering LDL cholesterol.