The possible effect of "white coat phenomenon" on blood pressure (BP) measured in epidemiologic field studies often is not considered, and practical alternatives to health care worker- or investigator-measured BP have not been evaluated extensively. In a study of occupational lead exposure and BP, investigators measured BP using a mercury sphygmomanometer, and study subjects measured their own BP using an oscillometric monitor (OMRON 705 CP, Omron Healthcare). Eighty-four subjects participated, although we eliminated two from further analysis when we found their data to be unreliable. Self-measured BP was positively correlated with investigator-measured BP (systolic: r=0.81, p=0.0001; diastolic: r=0.78, p=0.0001). Overall, self-measured BP was lower than investigator-measured BP. For systolic BP, the mean difference in paired measurements was 2.50 mm Hg +/- 9.97 (p=0.03) and for diastolic BP the mean difference was 2.26 mm Hg +/- 5.87 (p=0.0008). Twenty- seven (32%) and 23 (27%) subjects were classified as hypertensive based on the investigator and self measurements, respectively, according to published criteria for each type of measurement. Agreement between measurement methods was fair to good (kappa=0.55, 95% CI=0.35,0.74). Although the two approaches to measuring BP yielded differing results, it is unclear whether these differences would affect the results of epidemiologic studies. Further research is needed to clarify whether self-measured BP is a better BP measure and how it is related to known and suspected risk factors for hypertension.