BACKGROUND: Reincarceration in prison or jail correlates with non-sustained HIV viral suppression, but HIV treatment outcomes in released prisoners who are reincarcerated have not recently been systematically assessed despite advances in antiretroviral treatment (ART) potency, simplicity, and tolerability. METHODS: In a retrospective cohort of reincarcerated inmates with HIV in Connecticut (2005-12), we used longitudinally linked demographic, pharmacy, and laboratory databases to examine correlates of viral suppression. The primary outcome was viral suppression on reincarceration, defined as viral load lower than 400 RNA copies per mL. FINDINGS: Of 497 prisoners and jail detainees with HIV, with 934 reincarcerations, individuals were mostly unmarried, uninsured, and black men prescribed a protease-inhibitor-based ART regimen. During the median 329 days (IQR 179-621) between prison release and reincarceration, the proportion of incarceration periods with viral suppression decreased significantly from 52% to 31% (mean HIV-RNA increased by 0·4 log; p<0·0001), lower than Connecticut's HIV-infected prison population and those prescribed ART nationally. 158 (51%) of 307 individuals with viral suppression on release had viral suppression on reincarceration. Viral suppression on reincarceration was associated with increasing age (adjusted odds ratio [aOR] 1·04, 95% CI 1·01-1·07), being prescribed non-nucleoside reverse transcriptase inhibitor-based regimens (1·63, 1·14-2·34), and having higher levels of medical or psychiatric comorbidity (1·16, 1·03-1·30). INTERPRETATION: Identification of individuals most at risk for recidivism and loss of viral suppression might mitigate the risk that repeated reincarceration poses to systems of public health and safety. FUNDING: Bristol-Myers Squibb Virology, Patterson Trust, and National Institute on Drug Abuse.