A cross-sectional study of 90 bridge painters from 13 worksites and 8 contractors in Massachusetts and New Hampshire, was conducted over a two year period from 19941995. The aim of the study was to investigate the reasons for persistent elevated blood lead levels among bridge painters involved in lead abatement and bridge painting activities. Abatement activities ranged from complete deleading of the bridges and repainting to repainting previously deleaded bridges. Deleading methods observed included dry blasting, wet blasting, pressure washing, power tooling, and hand tooling. The worksites were evaluated for 14 days during which biological monitoring for blood lead and environmental sampling for ambient, surface and skin lead was conducted. Quantitative and qualitative information on the characteristics of the worksite, personal activities and hygiene, meteorological data, and lead abatement methods were gathered, including interviews with workers and contractors. This information was used to create indices of personal and site hygiene as well as other index variables that were used as exposure modifiers in epidemiologic models of blood lead levels. Blood lead testing results of interest include the following: (1) Blood lead levels did not differ significantly between the first and second year of the project. (2) Mean blood lead values increased significantly during the 14 day interval. (3) Mean blood lead levels differed by contractor, e.g., there were some "good" and "bad" contractors as reflected in the blood lead values. (4) The blood lead test distribution for the study participants was in excess of OSHA's estimated values based on implementation of the 1993 Construction Lead Standard. In 1994, 19% of all BL tests were in excess of 25 (gm/dl) which increased to 30% of all BL tests in 1995. There is no obvious explanation for this change. (5) Significant predictive factors in explanatory statistical models of the blood lead parameters include cumulative exposure, respirator wipe measurement, smoking on the job, training, use of personal protective equipment, months of bridge work, and years in construction. There is strong evidence to suggest a potential for bystander exposure and take-home lead. The mean levels of lead found in the clean side of the decontamination unit and the steering wheel of workers' automobiles is quite high, allowing the potential for take home lead through contaminated cars and possibly personal items stored in the decontamination unit. The mean exposure levels for samples taken from less than 20 feet from the containment ranged from 3 ugm/m3 to 132 ugm/m3, thus potentially exposing bystanders less than 20 feet from the containment to lead dust in excess of the PEL. A second finding of relevance is the wide range in the percent good practice between sites and between indices. The range in percent good practices suggests that there is room for improvement in personal behavior and site hygiene practices. Personal hygiene habits of the workers that were summarized into the personnel hygiene index had a mean percent good practice of a mere 22% in 1994, and improved to 66% in 1995. The same applies to respirator program index which was at 39% in 1994 and improved to 88% in 1995. Improvement in the personal hygiene practices of workers and the respirator program are an important means of reducing exposures. However, the frequency of respirator use and the type of respirators used by workers was not adequate to protect them against exposures in excess of the OSHA PEL. The data shows that in 1994, 34% of the time tasks were performed without the use of a respirators. That was reduced to 24% in 1995. The mean exposure levels were above the PEL for all the tasks except painting rails and assisting painters. Thus, some form of respirator use would be required to protect workers from exposures above the PEL while performing most of the tasks, yet many of the tasks were performed without the use of respirators. In many tasks, even when some form of PPE was used, it was not always of a type adequate to reduce workers' exposures to below the PEL. Finally, the data suggest that personal and site hygiene practices improved greatly from the first sampling year (1994) to the second year (1995). The hand wipes, respirator wipes, car steering wipes and decontamination unit wipes all improved in the second year (1995). Generally, in 1994 a greater percent of the wipe samples were in the high exposure category than in 1995 for all of wipe samples. A similar trend was also observed in the indices of site and personal hygiene which showed a remarkable improvement in all the indices in the second year. This improvement could be attributed to the impact of the new OSHA regulation of lead in construction. However, this conclusion must be drawn with caution as the improvement could also be attributed to the different lead removal methods used in the second year or low exposure activities on the day of sampling.