Aims/hypothesis The apolipoprotein B (ApoB):apolipoprotein A (ApoA)-I ratio may be a better indicator of cardiovascular disease (CVD) risk in people with type 2 diabetes than traditional lipid risk markers (LDL-cholesterol, HDLcholesterol and triacylglycerol), but whether the ApoB: ApoA-I ratio should be used to indicate lipid-lowering therapy is still debated. Methods The Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study randomised 9,795 patients with type 2 diabetes to fenofibrate (200 mg daily) or placebo and followed them up for a median of 5 years. We compared ApoB, ApoA-I, ApoAII and the ApoB:ApoA-I ratio with traditional lipid variables as predictors of CVD risk. We estimated the HR of the effect of 1 SD difference in baseline concentrations of lipids, apolipoproteins and respective ratios on the risk of CVD events and also used receiver operating characteristic curve analysis. Results In the placebo group, the variables best predicting CVD events were non-HDL-cholesterol:HDL-cholesterol, total cholesterol:HDL-cholesterol (HR 1.21, p<0.001 for both), ApoB:ApoA-I (HR 1.20, p<0.001), LDL-cholesterol:HDLcholesterol (HR 1.17, p<0.001), HDL-cholesterol (HR 0.84, p<0.001) and ApoA-I (HR 0.85, p<0.001). In the fenofibrate group, the first four predictors were very similar (but ApoB: ApoA-I was fourth), followed by non-HDL-cholesterol and ApoB. Lipid ratios and ApoB:ApoA-I performed better than any single lipid or apolipoprotein in predicting CVD risk. Conclusions/interpretation In patients with type 2 diabetes in the FIELD study, traditional lipid ratios were as strong as the ApoB:ApoA-I ratio in predicting CVD risk. The data provide little evidence for replacement of traditional lipids and their ratios with measures of ApoB, ApoA-I and their ratio.