Although statins reduce the risk of non-haemorrhagic strokes and transient ischaemic attacks (TIA), little is known about the efficacy of fibrates. This situation has been partly remedied by the recent publication of two-fibrate based trials--The Veterans Affairs High Density Lipoprotein Cholesterol Intervention Trial (VAHIT) and the Bezafibrate Infarction Prevention Trial (BIP). In BIP, bezafibrate did not significantly reduce the risk of a cerebrovascular event (CVE). Bezafibrate increased the high density lipoprotein cholesterol (HDL) level by 18% to 40 mg/dl (1.03 mmol/l) and decreased triglyceride (TG) levels by 21% to 115 mg/dl (1.29 mmol/l). In contrast, in VAHIT, gemfibrozil significantly reduced the risk of investigators designated stroke (P=0.04) and TIA (P<0.001). Gemfibrozil increased HDL by 6% to 33 mg/dl (0.85 mmol/l) and decreased TG by 31% to 110 mg/dl (1.25 mmol/l). However, the baseline low density lipoprotein cholesterol (LDL) levels were higher in BIP than in VAHIT (148 versus 111 mg/dl; 3.82 versus 2.87 mmol/l). LDL levels were not markedly altered by treatment in either trial. Fibrates can improve several CVE predictors, like fibrinogen, lipoprotein (a), insulin sensitivity and platelet activity. Furthermore, lowered HDL and/or raised TG levels are associated with an increased risk of a CVE; fibrates are an appropriate treatment for this lipid profile. In conclusion, the evidence suggests that not only total cholesterol and LDL, but also HDL and TG levels predict the risk of a CVE. Statins, fibrates or a combination of these drugs can modify these variables.