Interpretation of lung-function test results, specifically the forced vital capacity and forced expiratory volume in one second, generally involves the comparison of these parameters with reference values based on an individual's age, height, sex, and race. Such comparisons are often used to make important decisions concerning an individual, such as job placement or disability rating. Several studies have shown that predicted values for African Americans are approximately 15% less than those for Caucasians, most likely because of the use of standing height to estimate the size of the thorax. When an adjustment for race is applied to reference values based on a Caucasian population, a single value (15%) is usually applied to all individuals. When using a group of blue-collar workers (766 Caucasian and 633 African-American subjects) without any race adjustment, 10.2% of the Caucasians and 37.4% of the African-American subjects were below the lower limit of normal. When a single adjustment factor was used, 11.5% of the African-American subjects were below the lower limit of normal. Between-subject variability within an ethnic group was far greater than variability between groups. Our results suggest that although a difference between Caucasian and African-American test results for forced vital capacity and forced expiratory volume in one second exists, an application of a single adjustment factor universally applied to all individuals, regardless of their age, sex, and height, is not optimal, and alternative approaches are needed.