Background Evidence from animal studies suggests that minocycline may reduce lobar intracerebral hemorrhage (ICH) recurrence in cerebral amyloid angiopathy, possibly by inhibiting perivascular extracellular matrix degradation in cerebral small vessels. There is currently no evidence of its safety or efficacy in humans with cerebral amyloid angiopathy. Methods and Results To provide preliminary data to support future studies of minocycline's efficacy, the authors performed a retrospective single‐center cohort study to assess the incidence of recurrent ICH in patients with an aggressive clinical course of probable cerebral amyloid angiopathy who had been prescribed minocycline off‐label via shared decision‐making. Crude incidence rate ratios were calculated to compare incidence rates before versus after treatment. Sixteen patients (mean age at minocycline initiation, 66.3±3.5 years; women 62.5%; median of 3 lobar ICHs [range, 1–6]) were initiated on minocycline and followed for a median of 12.4 months (range, 1.8–61.4 months). Adverse events were reported in 4 of 16 patients (gastroenteric, n=3; dizziness, n=1) and were considered mild. ICH incidence sharply increased the year before minocycline initiation compared with the preceding years (2.18 [95% CI, 1.50–3.07] versus 0.40 [95% CI, 0.25–0.60] events per patient‐year) and fell to 0.46 (95% CI, 0.23–0.83) events per patient‐year afterwards. Incidence rate ratios of recurrent ICH after minocycline was lower (0.21 [95% CI, 0.11–0.42], P <0.0001) compared with the year before initiation. Conclusions Minocycline appeared safe and generally tolerated in a small group of patients with clinically aggressive cerebral amyloid angiopathy and was associated with reduced ICH recurrence. Determining whether this reduction represents a biological response to minocycline rather than a regression to the mean, however, will require a future controlled treatment trial.