Laboratory studies with competitive athletes are an important component of many sports science programs and are often designed to augment coaching. For ergometric studies to have any substantial value in understanding the conditioning of the athlete, they must correlate with performance and track longitudinal changes in fitness. Many laboratory tests satisfy the first criterion but fail in the second. This paper details historical experience of laboratory studies with one group of athletes, speed skaters, with the intent of demonstrating the difficulty in achieving a test that satisfies both criteria. On the basis of this experience, it is to be hoped that others charged with the scientific support of athletes and their coaches may arrive at satisfactory tests more rapidly.