In this study, respirable crystalline silica exposures to furnace relining workers of 7 exposure groups were assessed by conducting personal respirable dust samplings. All possible pollutant sources were identified for each exposure group through field observations, and bulk samples were randomly collected from each identified pollutant source. All collected samples were analyzed for their tridymite, cristobalite, and quartz contents by using the X-ray diffraction method. Results show that quartz was the only detectable crystalline silica content. We found that the resultant respirable quartz exposure levels presented in sequence for the 7 exposure groups (sand blasting > bottom ash cleaning > wall demolishing > relining > others > grid repairing > scaffold establishing) were different from that of the corresponding respirable dust exposure levels (bottom ash cleaning > wall demolishing > sand blasting > relining > grid repairing > scaffold establishing > others). 87.3-100% of workers' respirable quartz exposures of the 7 exposure groups exceeded the TLV-TWA (0.025 mg m(-3)) indicating appropriate control strategies should be taken immediately. By comparing the fractions of quartz contained in personal respirable dust samples with that contained in all possible pollutant sources for each exposure group, this study identified main pollutant sources for each exposure group as: bottom ash cleaning and scaffold establishing: bottom ash; sand blasting: blasting sand; wall demolishing: refractory cement + wall ash; wall relining: refractory brick; grid repairing: wall ash + refractory cement; grid repairing: wall ash + refractory cement; others: blasting sand + bottom ash. Finally, effective control strategies were proposed for exposure reduction by using above information together with our field observations. (C) 2007 Elsevier B.V. All rights reserved.