Background: Out of office blood pressure (BP) measurements, especially home BP monitoring, are recommended for diagnosis and management of hypertension by many societies including by the Canadian Hypertension Education Programme (CHEP). Though validation protocols exist for individual monitors, little data exists on the real world accuracy of home BP monitors, in actual use. We undertook this systematic review to estimate the accuracy of home BP monitors as reported in the literature. Methods: We undertook a literature search of MEDLINE and EMBASE from 1946 and 1947 until April 2015 respectively. We included studies that evaluated the accuracy of home BP devices against a mercury sphygmomanometer considered as the gold standard. Two reviewers independently selected studies, extracted data and assessed quality. Disagreements between the two reviewers were resolved by a third author. Results: Our search revealed 797 non-duplicate citations. After applying selection criteria, nineteen studies, involving 4954 patient-devices (median 91, interquartile range 69, 489) were included in the systematic review. The reported inaccuracy of home BP monitors compared to mercury sphygmomanometer ranged from 10% to 72% for systolic BP, with each study using different thresholds for inaccuracy. The absolute mean difference for systolic BP between home BP monitor and the standard ranged from 2.4 mm Hg to 10.4 mm Hg and for diastolic BP from 1 to 8.7 mm Hg. Conclusion: The existing literature reports a relatively high degree of inaccuracy in home BP monitors being used. Data is limited by varying definitions being used for reporting inaccuracy. Consideration should be given to standardised definitions of accuracy, and real world monitoring of accuracy as home BP monitors use for clinical decision making becomes widespread. Research into predictors of inaccuracy is also necessary.