While previous studies have investigated country-of-origin effect from various angles, it remained unexplored the extent to which Country of Origin Labeling (COOL) affects U.S. beef imports from specific countries. Using choice-experiment data, willingness to pay (WTP) for Australian, Canadian beef in addition to other enhancement attributes were estimated with a Mixed Logit Model and a Latent Class Model. The results revealed unobserved taste heterogeneity and important differences in the WTP between the imported and domestic steak. The Latent Class Model estimated the range of discount needed for consumers to switch from U.S. to Canadian steak was a range from $1.09 to $35.12 per pound. Results from the Mixed Logit Model reiterated strong domestic preference. Significant positive WTP for BSE-tested, traceable, and tenderness-assured beef were also observed. In addition, perceived risk theory was utilized to explain the difference in WTP for domestic and imported beef. The psychometric method proposed in Pennings et al. (2002) were adopted, which disentangled perceived risk into risk perception and risk attitude. Using a mixed logit model with error component specification, the result revealed a strong link between risk perception and risk attitude towards consumer choice of country-of-origin labeled beef. Specifically, we found that perceived risk factors have a stronger impact on imported beef than domestic beef, which could partially explain consumers’ aversion towards imported beef. Lastly, the perceived risk framework was expanded to explain variation in the WTP for traceable and BSE-tested beef. The results indicated significant and non-linear impact from risk attitude and risk perception to WTP for the attributes. In addition, BSE-concern, and perceived level of control agribusiness has on food safety significantly influenced WTP for traceable and BSE-tested beef.