Background: Measurements of the work ability subjective assessment, using the work ability index (WAI), are widely applied in the examination of workers. The measurement results suggest that the low level of work ability, which is determined by work-burden factors, health condition, and lifestyles of persons under study, can be a predictor of earlier retirement. The aim of the study was to find out whether WAI can be used in Polish conditions and to identify personal traits and/or job characteristics and conditions of its performance that generate the risk of low work ability. Materials and Methods: The cross-sectional study embraced 669 men and 536 women at the working age, representing different occupations and exposed to various factors. They self-assessed their work ability by completing a questionnaire that allows to determine WAI values. They also characterized their jobs in terms of physical burden, occupational stress, harmful and strenuous factors, work fatigue, chronic fatigue, and lifestyle. Based on the energy expenditure and health condition (number of diseases), the work burden was objectively defined. A model of multivariate logistic regression was used to assess the effect of the analyzed factors on the risk of low or moderate work ability. Results: The level of work ability in the study group was lower than that observed in analogous occupational groups in other European countries. The results of the analysis indicate that job characterizing factors and workers' individual traits exert a stronger effect on the level of WAI components, which reflect a subjective assessment of work abilities, than factors concerning health conditions. Highly stressogenic work and low tolerance of work burden as well as personal traits (age, frequent alcohol consumption among men and non-occupational burdens among women) represented risk factors responsible for low or moderate VAI values. Conclusions: The measurement of work ability index is an indirect assessment of workers' physical state, and it slightly depends on objective work burdens. Med Pr 2008;59(1):9-24