A unique discrete choice experiment (DCE) is used to estimate the relative importance of quality attributes to Australian beef consumers. In the DCE, consumers choose their preferred beef steaks from options varying in a large number of intrinsic (marbling and fat trim) and extrinsic/credence (brand, health, forage, meat standards/quality, and production and process claims) attributes. This study is the only known DCE to present these attributes to consumers visually – in a manner that more realistically simulates the retail choice scenario for beef and allows us to evaluate the relative importance of attributes that consumers use both consciously and unconsciously when making product choices. Respondents’ beef choices were analyzed using a latent class scale adjusted choice model. We address two import issues that have potentially strong implications for the validity of estimated attribute values: intrinsic attributes are likely to be underestimated in their importance if not presented visually; and DCEs that exclude important attributes (such as intrinsic characteristics) are likely to overestimate the value of product characteristics. The results suggest that visual attribute level presentation in DCEs results in less biased value estimates. Therefore, it is not only important to consider what attributes to include, but also how you present the attributes.