Dietary fatty acids (FA) are the major determinants of blood lipids, and measurements of plasma phospholipid FA (PL-FA) composition that reflect the dietary intake of FA may provide insights into the relationships between diet and CHD. We assessed CHD mortality associations with PL-FA (SFA, PUFA and MUFA) levels measured in a nested case–control study of 116 cases of CHD death and 239 controls that were frequency-matched for age and employment grade. The participants had plasma levels of total cholesterol, LDL-cholesterol (LDL-C), HDL-cholesterol, apo B and apo A1, C-reactive protein (CRP) and fibrinogen recorded. SFA levels were significantly positively correlated with total cholesterol, LDL-C, apo B, CRP protein and fibrinogen. By contrast, phospholipid-PUFA were inversely associated with CRP, but not with any of the lipids. A higher SFA content (top v. bottom quarter) was associated with a 2-fold higher risk of CHD (OR and 95 % CI: OR 2·12; 95 % CI: 1·13, 3·99), and an equivalent difference in PUFA was associated with a halving in CHD risk (OR 0·49; 95 % CI: 0·26, 0·94), but MUFA was unrelated to CHD risk. These associations were substantially attenuated, after additional adjustment for lipids and inflammatory markers. Higher levels of saturated fat and lower levels of polyunsaturated fats were each associated with a higher risk of CHD in elderly men, and these associations were partly explained by their effects on blood lipids and biomarkers of inflammation.