Background: Long-acting injectable cabotegravir is a novel integrase inhibitor currently in advanced clinical development for HIV prevention and treatment. We aimed to assess the terminal phase pharmacokinetics and safety of long-acting injectable cabotegravir in participants included in the HPTN 077 trial. Methods: HPTN 077 was a multicentre, double-blind, randomised, placebo-controlled phase 2a trial done at eight sites in Brazil, Malawi, South Africa, and the USA. Participants (aged 18-65 years), who were HIV-uninfected and at low-risk for HIV, were randomly assigned (3:1) to long-acting injectable cabotegravir (800 mg given three times at 12 week intervals or 600 mg given five times, administered at one 4 week interval, and every 8 weeks thereafter) or placebo. Participants were followed up to 76 weeks after final injection. In a prespecified analysis of secondary and exploratory outcomes, we assessed the safety, measured by the proportion of participants with grade 2 or worse adverse events, and pharmacokinetics, measured by apparent terminal phase half-life (t1/2app) and estimated time to lower limit of quantification (LLOQ) of long-acting injectable cabotegravir during the injection phase (defined as the time between first injection and 12 weeks or 8 weeks after the last injection in cohort 1 or cohort 2 respectively) and tail phase (defined as the time between final injection and 52-76 weeks post-final injection). Safety was analysed in all participants who received at least one injection. Pharmacokinetic analyses included all participants who had received at least one injection and had at least three cabotegravir measurements higher than the LLOQ after the final injection. Pharmacokinetic outcomes were estimated using non-compartmental methods. The trial is completed, and was registered with ClinicalTrials.gov, Findings: Between Feb 9, 2015, and May 27, 2016, 177 participants (134 participants in the cabotegravir group [74 participants in cohort 1; 60 participants in cohort 2] and 43 participants in the placebo group [25 participants in cohort 1; 18 participants in cohort 2) were enrolled and received at least one injection and thus were included in the safety analysis. The incidence of grade 2 or worse adverse events was significantly lower during the tail phase than the injection phase (p<0·0001). At 52-60 weeks after final injection, nine (23%) of 40 male participants had detectable cabotegravir concentrations and at week 76, four (13%) of 30 male participants had detectable cabotegravir concentrations compared with 52 (63%) of 82 female participants and 27 (42%) of 64 female participants at the same timepoints. The median time from the last injection to the time when cabotegravir concentration decreased below the LLOQ was 43·7 weeks (IQR 31·1-66·6; range 20·4-152·5) for male participants and 67·3 weeks (29·1-89·6; 17·7-225·5) for female participants (p=0·0003). t1/2app was longer for female participants than male participants (geometric mean fold-change 1·33, 95% CI 1·06-1·68; p=0·014), and longer for participants with a high body-mass index (BMI) than those with a low BMI (1·31, 1·06-1·63; p=0·015). Interpretation: The clinical significance of the long pharmacokinetic tail of cabotegravir observed in female participants compared with male participants, and those with higher BMI compared with a lower BMI, need to be addressed in future trials.