Efforts to characterize nonoccupational noise exposures have focused primarily on infrequent, episodic events. Few studies have assessed noise levels resulting from routine daily activities. In the current study, 112 construction workers wore datalogging noise dosimeters and simultaneously completed activity logs during two phases of data collection. The 81 subjects monitored in phase 1 received logs listing numerous preselected occupational and nonoccupational activities, while the 31 subjects monitored in phase 2 used free-field logs and reported nonoccupational activities in greater detail. Nearly all of the 221,439 1-min intervals of nonoccupational L(eq) level and activity reporting were below 70 dBA; only a small percentage exceeded 80 dBA. The primary contributor to nonoccupational noise exposure was traveling in a car or bus, while time at home contributed the least. One hundred seventy 24-h L(eq) levels were computed from the 1-min noise level data. The percentage of phase 2 workday L(eq(24)) levels which exceeded 80 dBA was higher than that of the nonworkday levels. The mean L(eq(24)) level of phase 2 workdays was higher than that of nonworkdays, and the difference was statistically significant. Routine nonoccupational noise exposures contributed much less to total noise dose than occupational exposures in the subjects evaluated.