Objectives. —To measure occupational exposures to environmental tobacco smoke in diverse settings, including offices and production areas, and to evaluate the effectiveness of policies that restrict or ban smoking in the workplace. Design. —Survey. The average weekly concentration of environmental tobacco smoke was measured with passive monitors that sample nicotine. Approximately 25 samplers were placed in each worksite for 1 week. Setting. —Twenty-five Massachusetts worksites, including fire stations, newspaper publishers, textile dyeing plants, and manufacturers of valves, fiberoptics, flight instruments, batteries, adhesives, semiconductor equipment, filters, and tools and dies. Samples were collected in offices and production areas. Main Outcome Measures. —The distribution of nicotine concentrations in various work settings as a function of company smoking policy. These data were interpreted with three approaches: comparing measured concentrations with a published risk assessment; comparing occupational exposures with home exposures; and evaluating the "cigarette equivalents" to which workers were exposed. Results. —Worksite smoking policy had a major effect on the nicotine concentrations, which fell from a median of 8.6 μg/m3in the open offices at worksites that allowed smoking to 1.3 μg/m3in sites that restricted smoking, and to 0.3 μg/m3in worksites that banned smoking. The nonoffice workspaces were affected similarly, with median concentrations of 2.3,0.7, and 0.2 μg/m3at worksites that allowed, restricted, and banned smoking, respectively. Conclusion. —All three evaluation methods indicated that occupational exposure to environmental tobacco smoke presents a substantial risk to workers in the absence of a policy restricting or banning smoking. (JAMA. 1995;274:956-960)