Several findings of the author\'s own mental workload research are reviewed. Especially, factors, except task demand, influencing mental workload indexes are discussed. First, an experiment using the National Aeronautics and Space Administration Task Load Index (NASA TLX) is reviewed and the effects of a prior task load on a subsequent subjective workload are described. In a physiological measures section, the low reliability of the heart rate variability (HRV) parameters when the subjects\' respiration pattern are highly irregular, and the existence of low sweat respondent, although perspiration seems to be sensitive to the workload, are indicated. Furthermore, results of the task specific physiological responses and of a HRV parameter affected by the individual characteristics, like type A behavior pattern, are shown. Discrepancy between autonomic nervous system activity measures may be solvable by introducing a new concept of autonomic space (synergism of sympathetic and parasympathetic nervous system). Finally, the need for strategies of data management concerning the individual differences is emphasized.