We read with great interest the recent article by Baste et al in which the authors found that Norwegian hairdressers who were smokers or ex-smokers had a higher risk of infertility compared with hairdressers who were never smokers and control women working in nonhairdresser occupations. We also read with great interest the recent Letter to the Editor from Axmon and Rylander in which the authors commented on the Baste et al study. Interestingly, Axmon and Rylander conducted similar analyses to Baste et al in a Swedish population of hairdressers and controls and did not find an interaction between smoking and working as a hairdresser with respect to similar reproductive health outcomes. In response to the Letter from Axmon and Rylander, Baste and Moen suggested that the different findings between their study and that of Axmon and Rylander may be due to differences in smoking prevalence and age between the Norwegian and Swedish study samples. Further, Baste and Moen suggested that others collect or analyze data on smoking and reproductive outcomes in hairdressers, particularly in women older than 40 years of age. Recently, we conducted a study on Reproductive Outcomes in Salon Employees (ROSE) in the US. We previously published detailed methods for the ROSE study. Briefly, we collected data on reproductive health from 450 hairdressers and 511 women employed in other occupations (ie, teachers, nurses, real estate agents, sales clerks) using a 51-page survey that was mailed to registered female hairdressers and nonhairdressers aged 21 to 55 years in the Baltimore, Maryland area and its surrounding counties. Names and addresses of potential participants in the selected occupations and age range were obtained from a commercial mailing house, and recruitment proceeded via a mass mailing. To be included in the study, women had to have an intact uterus and both ovaries. All hairdressers were confirmed to be employed as hairdressers by their responses to detailed work history questions. Per the suggestion of Baste and Moen, we analyzed data obtained from the mailed surveys to determine if, in our sample, there was an interaction between smoking and working as a hairdresser with respect to miscarriage and infertility. For the miscarriage outcome, each reported pregnancy was treated as the unit of observation, and occupation (hairdresser/other) and smoking status (yes/no) at the time of each pregnancy were analyzed using repeated measures analyses of variance procedures. For the infertility outcome, each individual woman was treated as the unit of observation; infertility was determined using the question "Have you ever tried to become pregnant over a period of at least a year without success?" Logistic regression procedures adjusting for age at survey, body mass index, race, marital status, educational level, and current alcohol drinking were conducted to examine the association between being a hairdresser at the time of survey and infertility. There was not a statistically significant interaction between being a hairdresser and smoking for either the miscarriage or infertility outcome among all women in the study and among only women aged 40 years and older at the time of the survey. As reported in a previously published ROSE article, the odds ratio (OR) of miscarriage was significantly higher among women who smoked at the time of their pregnancy compared with women who did not smoke (all women: OR, 1.62; 95% CI, 1.11 to 2.38). Further, the OR of miscarriage were highest among hairdressers who smoked compared with the other study groups (nonhairdressers who did not smoke during pregnancy was the reference category). In contrast, smoking was not related to infertility in our sample nor was being a hairdresser related to either adverse reproductive outcome. As suggested by Baste and Moen, it is likely that different study population characteristics account for differences in results. Both the Baste et al and Axmon and Rylander studies used European study populations. Our study was based on a US study population. Further, the reproductive outcomes that were assessed in our study differed from those assessed in the other studies. We investigated miscarriage and infertility, whereas Baste et al and Axmon and Rylander examined delayed conception and spontaneous abortion. Given these study differences, it is not possible to definitively determine whether there is an interaction between smoking and working as a hairdresser with respect to all reproductive outcomes. Thus, we hope that others will collect and analyze their data as well help to determine whether an interaction exists between smoking and reproductive outcomes in hairdressers.