In the February 2013 issue of Current Oncology, Dr. Steven Narod critiqued the 2012 study by Brophy et al., which found elevated breast cancer risk among several occupational groups. He also expressed skepticism concerning the role that environmental exposures to industrial chemicals and pollution play in the causation of breast cancer. That skepticism reflects an outdated epidemiologic analysis undertaken 30 years ago. In 2010, the President's Cancer Panel in the United States drew attention to "the unacceptable burden of cancer resulting from environmental and occupational exposures," which it said was "grossly underestimated." The Panel stressed the need "to identify the many existing but unrecognized environmental carcinogens and eliminate those that are known from our workplaces, schools, and homes." In Europe, attention has been drawn to the need to identify "hidden" groups whose occupational exposure to carcinogens is underrepresented in intervention strategies (European Agency for Safety and Health at Work. Workshop on Carcinogens and Work-Related Cancer. 2012). Workers as a group tend to be more highly exposed to carcinogens, but they have been relatively neglected by epidemiologists. This lack of scientific attention could be characterized as a reflection of sex and class bias. The Brophy et al. study was conducted in response to those gaps. This comprehensive case-control study was considerably more than what Dr. Narod characterized as "an interesting pilot." It was a hypothesis-driven study based on findings from two previous studies by the authors. Dr. Narod's critique of the study design was also ill-informed. The authors acknowledge that there were social class differences between the cases and controls, just as there were exposure differences. That was precisely why the analysis carefully controlled for family income and education levels. Given the inevitable exposure misclassification, low-income (and generally more highly exposed) workers might be expected to show more residual excesses not accounted for in the exposure classification scheme used-as was observed. If participation rates were higher among non-cases in the service sector, as suggested by Dr. Narod, there should have been a nonspecific general excess across all manufacturing sectors-which was not observed. Most minor non-plastics sectors had deficits similar to those observed in services (liquor, tobacco, wood, printing, petroleum, glass and ceramic, electrical, jewelry and furniture) and in transportation. Differences in participation rates of controls would not generate higher estimated relative risks when cases are restricted by receptor type or menopausal status, as was observed.