Although many studies have been made of hypertensive left ventricular (LV) hypertrophy regression, most have been marred by small numbers of subjects, unrepresentative populations, short study duration, lack of comparison between agents, unblinded echocardiographic readings and inappropriate statistical methods. Meta-analysis and critical reviews of this flawed literature suggest, but do not prove, that some antihypertensive drug classes may reverse LV hypertrophy more effectively than others. Eagerly awaited multicentre trials have not provided definitive answers because of pre-emptive success of concomitant nonpharmacological therapy in reversing both hypertension and LV hypertrophy or excessive drop-out of study participants. Future studies that avoid the above limitations are needed to address three sets of questions: relatively small ones (n > or = 40 to 60) to investigate pathophysiology in detail or explore the effects of new agents; medium-sized ones (n > or = 300-400, time > or = 1 year) to determine definitively whether inter-agent differences in reduction of LV mass exist; and large long-term trials (n > or = 1200, time > or = 4 years) to determine whether LV hypertrophy reversal improves prognosis over and above blood pressure reduction and the type of treatment used.