Fatigue, a common symptom of multiple sclerosis (MS), is associated with impairment in performing daily activities, poor quality of life and premature retirement from the workforce. There is little doubt that patients with MS can exhibit marked weakness and other objective signs of physical fatigue, but whether cognitive performance by patients declines more rapidly than that by controls as a function of time engaged in mental activity remains controversial. Krupp and Elkins (2000) reported more rapid deterioration of performance by MS patients on two of five cognitive measures when subjects performed 3 hr of continuously effortful tasks between baseline and posttest. Others, using shorter and less taxing interpolated tasks found similar changes in performance over time for patients and controls. In the present study participants completed a timed walk, fatigue ratings and four cognitive tests that emphasized processing novel information before they went to work and again after completing a normal workday. Patients with MS walked somewhat more slowly and performed more poorly on two of the cognitive tests, but they did not show more decline from baseline to posttest on any of these objective measures of cognitive fatigue. By contrast, subjective ratings of fatigue showed a greater increase over the day for patients than for controls. These results confirm other reports that patients' subjective ratings of their fatigue are not valid indicators of their actual performance on cognitive tests. Furthermore, laboratory studies that report "objective" cognitive fatigue in MS may utilize conditions that model the cognitive fatigue associated with the jobs patients actually perform rather poorly.