Ballard and Cream, henceforth referred to as the “noes”, and Shah and Suh (the “ayes”), express differing views on the matter of prescribing atypical antipsychotics to patients with dementia. They provide new and useful information which demonstrates that the psychogeriatric community is striving conscientiously to improve the management of BPSD. Despite some strong words, the “noes” ultimately admit that “long-term treatment with neuroleptics should (only) be considered in extreme circumstances and with the informed consent of next of kin”. Although this conflicts with their assertion that “there is clearly no rational reason for prescribing…(neuroleptics),” it suggests that the debate is really about defining where the threshold for prescribing neuroleptics to people with dementia should lie, rather than whether we should or shouldn't use the drugs at all. The “ayes” proffer the liberal opinion that clinicians should offer details about efficacy and side-effects to patients and carers in order that they might “weigh up carefully the risks and benefits” of treatment. It is not said in what proportion of cases with BPSD this avenue should be taken, nor is it clear how the final decision will be reached. This is too simplistic a solution. We cannot deal with this issue just by passing the buck to the patient and carer.