Termination of pregnancy (TOP), or the more common and confusing name in the USabortionusually preceded by adjectives such as 'elective' or 'therapeutic,' is a procedure fraught with societal controversy and conflict. The fuel for this conflagration arises not from biology or epidemiology, but on differences about what status the embryo or fetus should be assigned vis-a-vis the moral status of his or her mother, specifically her autonomy. 1 Perinatal epidemiology lacks the tools to answer this metaphysical question, which can only be informed by culture, ethics, and religion.In the mid 1980's in the US, proponents of restricting access to TOP tried to use epidemiology to support their arguments for restriction. They embraced utilitarianism, and focused on a possible, and controversial, association between TOP and breast cancer later in a women's life. 2 If the intermittently detectable association between TOP and breast cancer is actually causal, these restriction advocates made an epidemiologic case against TOP using scientific materialism rather than metaphysics. This violates the limitations of epidemiology stated in the first paragraph and looks to epidemiology to answer controversies that science is incapable of responding to. It is an example of the proverbial 'seeking a loaf of bread from a hardware store.' If one ascribes a moral status to embryos or fetuses (as I do), then even if TOP prevented subsequent breast cancer, this utilitarian argument would not override ethical conclusions and responses.What then can epidemiology do? Our discipline can carefully look at associations between exposure and outcomes and make informed predictions about the likelihood of causality. TOP stretches epidemiology to its limits for reasons directly tied to the aforementioned societal controversy swirling around it. First, due to the ethical principles inherent in a woman's decision to exercise her autonomy and undergo TOP, exposure to the procedure cannot be randomly assigned. Thus, there will never be randomised trials and all "evidence" about safety and harms will be derived from observational studies. A second consequence of the ethical disagreements is a hesitancy for individuals to report prior TOP, leading to inherent biases about exposure that make untangling relationships even more difficult. 3 KC and colleagues in Finland provide us a great service in relying on government registries where both TOP exposure and subsequent outcomes are reliably recorded and can be linked. 4 Registry linkage overcomes the problems of bias arising from TOP exposure ascertainment. They explore the important outcome of subsequent preterm birth. They report associations between surgical TOP and small for gestational age and extremely preterm birth. The latter association was only seen with repetitive surgical TOP. 4 These findings confirm the conclusions of two meta-analyses 5,6 and are biologically plausible via a mechanistic path of damage to the cervix during manual dilation.While the associations reported are small and limited by p...