The growing cost of environmental regulation has fueled a debate over cost-effectiveness in improving environmental quality. Yet costs of regulation are difficult to measure, since abatement costs should properly include productivity losses or gains associated with abatement technologies. Technological innovations are often both productivity enhancing and emission reducing so an economic assessment of costs imposed by regulation requires a comparison with a counterfactual in which plants adopt technologies (that may well be emission reducing) in the absence of regulation. Similarly, the benefits of regulation are also difficult to measure, since that analysis involves an unobserved counterfactual, the change in air quality in the absence of the regulation. This study estimates the costs and benefits of air quality regulation in manufacturing using an approach that respects the notion of a counterfactual, using comparison plants and regions subject to less stringent regulations to estimate costs and benefits of local air quality regulation. The study requires construction and documentation of new panel data on local regulations in California. It will develop and document a previously unused panel on plant emissions. These will be linked to existing panel data on abatement and production by plants from the Census Bureau. This approach requires developing econometric methods for constructing micro-regulatory data in a way that allows subjective coding decisions without biasing results. The decentralized, multi-tiered structure of U.S. environmental regulation gives rise to rich variation across geography and industries. The study exploits that variation to estimate the effects of regulation on productivity, emissions and air quality. Precision is achieved by combining 1) the Census Bureau national sample of plants, including abatement costs; 2) data on local regulation for three California air basins; 3) a Californian sample on plant emissions; and 4) monitoring station data on local ambient air quality in California. Linked micro data developed here on regulation, productivity emissions, and air quality will provide basic public goods for future research in environmental economics, productivity, public finance, industrial organization, and labor studies. It should allow further research on the cost-effectiveness of particular regulations. The project also contributes a useful methodology in the econometrics of micro-regulation.