The paper by Perrine et al (this issue, pp 150-4) is one in a series of recent reports of cognitive outcome after unilateral pallidotomy for patients with Parkinson’s disease.1-4 These studies have taken advantage of sophisticated and selective surgical techniques (and outcome measures) unavailable to earlier, poorly controlled, studies in the pre-levodopa era. More reliable targeting of globus pallidus interna lesions is likely to have reduced cognitive complications, although potentially, the margin for error remains small and there has been at least one report of hemisphere specific postoperative cognitive changes.4 Nevertheless, despite a considerable degree of individual variability (postoperative gains and deficits), experienced centres have not generally found clinically significant cognitive sequelae (in terms of statistically significant …