In previous studies, investigators have reported reduced mortality among women undergoing assisted reproductive technology (ART) treatment, possibly related to selection of healthy women into ART treatment. Our aim in this study was to explore the impact of relevant selection factors on the association between ART treatment and mortality and to explore effect modification by parity. Women treated with ART in fertility clinics in Denmark during 1994-2009 (n = 42,897) were age-matched with untreated women from the background population (n = 204,514) and followed until December 31, 2010. With adjustment for relevant confounders, the risk of death was lower among ART-treated women during the first 2 years after ART treatment (hazard ratio (HR) = 0.68, 95% confidence interval (CI): 0.63, 0.74), but there was no apparent difference after 10 years (HR = 0.92, 95% CI: 0.79, 1.07). Having children prior to ART treatment was associated with markedly reduced mortality (HR = 0.45, 95% CI: 0.38, 0.53), possibly due to better health among fertile women. While the frequencies of previous medical and psychiatric diagnoses among ART-treated and untreated women were similar, differences in disease severity could explain the reduced mortality among ART-treated women, as poor prognosis would make initiation of ART treatment unlikely. The survival advantage among ART-treated women is likely a selection phenomenon rather than a biological phenomenon.