We conducted a retrospective study of a set of previously published electrocardiographic data to investigate the possible direct association between levels of particulate air pollution and changes in ventricular repolarization -- the cardiac electrophysiologic process that manifests itself as the T wave* of the electrocardiogram (ECG) and that is definitively linked to and responsible for increased arrhythmogenesis. The published findings from this data set demonstrated a clear cardiac effect, namely, a reduction in heart rate variability (HRV) parameter values with increased levels of particulate air pollution (Pope et al. 2004), suggesting possible arrhythmogenic effects. Given this positive finding and the well-established sensitivity of cardiac repolarization to physiologic, pharmacologic, and neurologic interventions, and in light of emerging novel tools for assessing repolarization, we hypothesized that high levels of particulate air pollution would alter repolarization independent of changes in heart rate and, consequently, would increase arrhythmogenic risk. The likely mechanism of any deleterious effects on repolarization would be alteration of sodium, calcium, and potassium channels. The channel's structure, function, and kinetics are responsible for generating the cellular action potentials, which, when summed over the entire heart, result in the waves recorded by the ECG. A positive finding would provide evidence that increased levels of air pollution may be directly linked to increases in arrhythmogenic risk and, potentially, sudden cardiac death. The study population consisted of 88 nonsmoking, elderly subjects in whom multiple, continuous, 24-hour, 2-channel ECG recordings were collected, along with blood samples to evaluate inflammatory mechanisms (not pursued in the current study). The concentration of fine particulate matter (PM2.5, particulate matter with an aerodynamic diameter < or = 2.5 microm) in daily samples was measured or estimated and used to trigger recording sessions for days considered to have "low" or "high" PM2.5 concentrations. Each subject participated in one to five recordings over the study period, and all subjects lived within the greater Salt Lake Valley in Utah. We reanalyzed these recordings using custom software that incorporated a magnitude function of the ECG -- the root mean square of all recorded leads (RMS ECG) -- to determine the following for each beat in the 24-hour recording: cycle length (RR); RR dispersion; the interval between the RMS R- and T-wave peaks (RT), a robust estimate of mean duration of ventricular action potential; the width of the RMS T wave (TW), a robust estimate of the range of repolarization times that relates to repolarization dispersion and arrhythmogenesis; the RMS QT interval (QT) measured from the QRS onset to T-wave offset of the RMS ECG; and the regression slopes of RT versus RR, QT versus RR, and TW versus RR, which provide estimates of so-called repolarization restitution, or rate dependency of repolarization, which also is associated with arrhythmogenesis. The study findings did not support the original hypothesis and demonstrated a lack of sensitivity of repolarization to changes in PM2.5 concentrations. None of the repolarization variables showed a statistically significant change between days of low and high PM2.5 concentrations, although we observed statistically significant differences for some variables using fixed-effects modeling. However, we did find a significant decrease in the standard deviation of cycle length, in concert with findings in the original study that showed a decrease in HRV parameter values. There was a slight but statistically insignificant increase in the width of the TW between recordings from days of low and days of high PM2.5, suggesting that, in a setting of prolonged exposure to high levels of PM, the original hypothesis might be supported. We conclude that in this study the short-term (day-today) differences in air pollution, specifically PM2.5 concentration, did not affect ventricular repolarization. A likely explanation for the negative result is that the day-today variability of repolarization (arising from autonomic influences, activity, and heart rate) far outweighs the changes that might be induced by air pollution, if any. In addition, the study may have been underpowered. The findings do not refute the possibility of the deleterious repolarization effects of PM, particularly over prolonged periods of exposure, but suggest the need for exposure studies that provide better controls. In light of recent studies, it is also likely that in an at-risk population -- for example, patients compromised with heart disease -- repolarization changes may be more apparent.