Background: No drug treatments are currently licensed for the treatment of borderline personality disorder (BPD). Despite this, people with this condition are frequently prescribed psychotropic medications and often with considerable polypharmacy. Preliminary studies have indicated that mood stabilisers may be of benefit to people with BPD. Objective: To examine the clinical effectiveness and cost-effectiveness of lamotrigine for people with BPD. Design: A two-arm, double-blind, placebo-controlled individually randomised trial of lamotrigine versus placebo. Participants were randomised via an independent and remote web-based service using permuted blocks and stratified by study centre, the severity of personality disorder and the extent of hypomanic symptoms. Setting: Secondary care NHS mental health services in six centres in England. Participants: Potential participants had to be aged >= 18 years, meet diagnostic criteria for BPD and provide written informed consent. We excluded people with coexisting psychosis or bipolar affective disorder, those already taking a mood stabiliser, those who spoke insufficient English to complete the baseline assessment and women who were pregnant or contemplating becoming pregnant. Interventions: Up to 200 mg of lamotrigine per day or an inert placebo. Women taking combined oral contraceptives were prescribed up to 400 mg of trial medication per day. Main outcome measures: Outcomes were assessed at 12, 24 and 52 weeks after randomisation. The primary outcome was the total score on the Zanarini Rating Scale for Borderline Personality Disorder (ZAN-BPD) at 52 weeks. The secondary outcomes were depressive symptoms, deliberate self-harm, social functioning, health-related quality of life, resource use and costs, side effects of treatment and adverse events. Higher scores on all measures indicate poorer outcomes. Results: Between July 2013 and October 2015 we randomised 276 participants, of whom 195 (70.6%) were followed up 52 weeks later. At 52 weeks, 49 (36%) of those participants prescribed lamotrigine and 58 (42%) of those prescribed placebo were taking it. At 52 weeks, the mean total ZAN-BPD score was 11.3 [standard deviation (SD) 6.6] among those participants randomised to lamotrigine and 11.5 (SD 7.7) among those participants randomised to placebo (adjusted mean difference 0.1, 95% CI-1.8 to 2.0; p = 0.91). No statistically significant differences in secondary outcomes were seen at any time. Adjusted costs of direct care for those prescribed lamotrigine were similar to those prescribed placebo. Limitations: Levels of adherence in this pragmatic trial were low, but greater adherence was not associated with better mental health. Conclusions: The addition of lamotrigine to the usual care of people with BPD was not found to be clinically effective or provide a cost-effective use of resources. Future work: Future research into the treatment of BPD should focus on improving the evidence base for the clinical effectiveness and cost-effectiveness of non-pharmacological treatments to help policy-makers make better decisions about investing in specialist treatment services.