The purpose of this paper is to present the assessment of magnetic field exposure conducted as a part of a nested case-control investigation of leukemia mortality in telephone lineworkers. For the purposes of exposure classification, telephone company jobs were initially divided into two classes: those with potential for working in an electric environment, referred to as linework jobs, and those not working in an electric environment, referred to as nonlinework jobs. Linework jobs were further divided into the following four categories: outside plant technicians (OPT), installation/maintenance/repair (IMR) technicians, central office technicians (COT), and cable splicing technicians (CST). These job groupings were based on similarity of work tasks and exposure environments. Emdex data-logging dosimeters were used to measure personal exposures to ELF magnetic fields for 204 telephone company workers. Three general classes of exposure indices were calculated for each exposure record: measures of central tendency, measures of peak or maximum exposure, and measures of exposure variability. CSTs had the highest full-shift mean and median exposure, 4.3 and 3.2 mG, respectively. CSTs also ranked the highest, with average peak, average 95th percentile, and average time above background equal to 99.2 mG, 11.1 mG, and 156 min, respectively. In addition, the results suggest the OPT and IMR technicians have exposures similar to nonlineworkers. Exposure classifications, therefore, which misclassify all lineworkers into one "telephone lineworker" job grouping are not appropriate and future studies should concentrate on cable splicing technicians.