Can heroin addicts give consent to research on trials in which heroin is prescribed to them? Analyses of addicts and informed consent have been objects of debate in several articles. Informed consent requires the agent not only to be competent but also to give consent voluntarily. This has been questioned because of alleged features of heroin addiction. Until recently the discussion has focused on heroin addicts' desires for heroin, whether these are irresistible and thus pose a problem for giving consent. Still, in light of empirical evidence, there seems to be a consensus more or less that the problem is not whether the addicts can resist their desire for heroin. A recent article concentrates specifically on heroin addicts' false assumptions of options and voluntariness. We argue that the prevailing framing of the options in this discussion in terms of heroin and access to it is problematic. The way in which the options are typically laid out suggests an assumption that participation in the research is allegedly based on the addicts' views on using the drug. We argue that this way of presenting the options is, first, a mismatch to the studies carried out and, second, symptomatic of potential misconceptions about heroin addiction and addicts. Furthermore, we also suggest that the account of voluntariness needs to be realistic in order for subjects to be able to give consent voluntarily in actual situations, and for medical research to carry out studies on improving outcomes in addiction treatment in an ethical way.