Although the ecologic effects of acid rain have been widely reported, relatively little is known about the effects of acidic air pollution on human health. Some epidemiologic and animal studies suggest, however, that acidity is an important determinant of the respiratory health effects of aerosols. The paper reviews some of that evidence and discusses its implications for the design and analysis of epidemiologic studies. Two types of exposure patterns are contrasted: peak exposures associated with air pollution episodes, and chronic exposures resulting from persistently high levels of air pollutants. Recent work on the analysis of repeated categorical outcome variables provides new methods for the analysis of episode studies. Studies of long-term exposure require comparisons among population groups, and these comparisons can be subject to the design effects characteristic of multistage sample surveys. The paper examines the implications of these design effects for epidemiologic studies. The paper also discusses the measurement errors induced by the use of outdoor measurements to quantify personal exposure, and ways to assess effects of such errors on conventional analyses of air pollution studies.