The color of beef longissimus dorsi muscle of a homogeneous group of fifteen males of Pirenaica breed in terms of the CIELAB color space was studied using the actual illuminating conditions existing in a representative fraction of retail displays operating in 51 local butcher's shops and supermarkets. All retail displays mounted fluorescent lamps and up to nine fluorescent models with different chromatic characteristics were found. Meat color was analyzed at four different times: immediately after cutting and after I day, 5 days, and 7 days of exposure to air. At each measurement day computed a*, b*, C* and hub color coordinates gave significantly higher values than the corresponding D65 values. All fluorescent models enhanced the red color of meat but none of them masked or enhanced meat color differences between samples or between measurement days. Significant differences in color coordinates were found between retail displays that mounted identical fluorescent models, proving the important role of uncontrolled environmental factors like daylight or scattered light from the walls and ceiling. With respect to computed color coordinates, the intrinsic natural variations in the color of meat samples were more important than variations in the illuminating conditions. (c) 2005 Wiley Periodicals, Inc.