To expand on the limited size and scope of construction silica exposure studies, a silica monitoring data compilation project was initiated through the American Conference of Governmental Industrial Hygienists Construction Committee. Personal silica exposure monitoring data was collected and analyzed from 13 private, research, and regulatory groups. An effort was made to collect as much detail as possible about task, tool, and environmental and control conditions so as much information as possible could be garnered. There were considerable data gaps, particularly with regulatory agency data, that represented over half of the data set. There were 1374 personal quartz samples reported with a geometric mean of 0.13 mg/m(3) and a GSD of 5.9. Descriptive statistics are reported by trade, task, tool, and data source type. Highest exposures were for abrasive blusters, surface and tuckpoint grinders, jackhammers, and rock drills. The sample period was important, with short-term samples (up to 2 hours) having considerably higher levels than midterm (2-6 hours) or longer (over 6 hours) samples. For nearly all exposure variables, a large portion of variable categories were at or over the quartz occupational exposure limit of 0.05 mg/m(3), including 8 of 8 trade, 13 of 16 task, and 12 of 16 tool categories. The respiratory protection commonly used on construction sites is often inadequate for the exposures encountered. The data variability within task and tool was very large, with some very high exposures reported for a broad spectrum of tools. Further understanding of the conditions leading to high exposures will require more detailed documentation of the sample characteristics following database design recommendations or systematic surveys of exposure in this complex industry.