Prospective evidence on the extent to which serum lipid concentrations in older persons respond to dietary modification is scarce. It is not clear whether such behavioral changes are relevant in the context of more commonly initiated treatments with lipid-lowering drugs. We therefore examined whether individual changes in the consumption of dietary fatty acids or main food sources were associated with changes in the serum lipid profile of older Australians. A total of 903 participants (> or =49 y) in the Blue Mountains Eye Study had complete data on fasting lipids and dietary intake from a validated FFQ at baseline (1992-1994) and 5- and 10-y follow-up examinations. Decreasing consumption of SFA and butter during the 10-y period were associated with moderate decreases in serum total cholesterol independently of initiation of lipid-lowering drug treatment [adjusted estimates were 0.018 +/- 0.007 mmol/(L x % energy (%en) from SFA (P = 0.01) and 0.055 +/- 0.015 mmol/(L x 5 g butter) (P = 0.0003), respectively]. Increased consumption of (n-3) fatty acids and fish was independently related to modest increases in serum HDL-cholesterol [0.067 +/- 0.026 mmol/(L x %en from (n-3) fatty acids) (P = 0.01) and 0.010 +/- 0.004 mmol/(L x 20 g fish) (P = 0.02)] and decreases in log-transformed serum triglyceride concentrations [P = 0.02 for (n-3) fatty acids and P = 0.02 for fish intake]. Hence, 10-y changes in the intake of dietary fatty acids and their food sources appear to have contributed to concurrent improvements in the serum lipid profile of older Australians, independent of concomitantly initiated lipid-lowering drug treatment.