We echo Waugh and colleagues' call for further studies to evaluate home blood pressure monitoring more fully. The main remaining issue is safety, which is why our group has developed a protocol for a trial to test the hypothesis that home based self-monitoring combined with a reduced schedule of antenatal visits, will result in less severe hypertensive disease at presentation to hospital. We do not of course suggest that monitoring of any sort actually reduces severe disease, simply that it might allow it to be detected at an earlier stage. Even this will be a major exercise. We estimate that 10,000 women will need to be involved. The issue of where the threshold for self-referral should be set, at 140/90 mmHg or 135/85 mmHg or even at some other level is vexed. Our preferred option is to use a 140/90 mmHg threshold, and to monitor this by reporting continuously all women with hypertension to an independent safety committee. This will not preclude later modelling studies to evaluate the effect of different thresholds on early detection and false positive referral rates. This protocol is currently under consideration by the UK Medical Research Council.