The aim of this study was to investigate the reliability and validity of the Fatigue Impact Scale (FIS) among patients with chronic obstructive pulmonary disease (COPD) and shorten the questionnaire. The empirically developed FIS, which comprised three subscales (cognitive, physical and psychosocial), was tested originally on Pipers' theoretical framework of subjective manifestations of fatigue, including behavioural, physical, emotional and cognitive expressions. The data analysed here consisted of responses from 296 patients with COPD who reported fatigue. The dimensionality of the FIS was examined using confirmatory factor analysis. A reduction of 15 items from the original FIS was made based on theory, modification indices and factor loadings. The results indicate that a nested-factor model with one general behavioural factor and three specific factors, physical, emotional and cognitive, shows acceptable fit. A modified version of 25 items, FIS-25 was developed. The original FIS and the FIS-25 were able to discriminate between patients with differing duration of fatigue. Test - retest correlations ranged from .70 to .85 for items and .94 for the total scale. Due to modification, the FIS-25 needs to be validated on a new group of patients with COPD.