BP readings that are obtained in the dialysis unit are commonly used to make therapeutic decisions by clinicians and to predict morbidity and mortality by epidemiologists. Dialysis unit BP are also incorporated in the recent guidelines to target BP control. The magnitude of the difference, overestimation or underestimation, and agreement between dialysis unit BP and ambulatory BP (ABP) are unknown. Articles were selected from Medline to identify those that reported both ABP and dialysis unit BP in hemodialysis patients. Bias was calculated as the difference between dialysis unit and the corresponding ABP. Agreement limits between the BP measurement techniques were assessed by pooled SD of the difference using Bland-Altman methods. Predialysis systolic BP generally overestimated ABP by a variable amount. The heterogeneity between BP measurements did not allow for pooling of the estimates. The agreement limits between the two BP was 41.7 to -25.2 mmHg. Predialysis diastolic BP also generally overestimated the ABP with wide agreement limits (23.7 to -18.9 mmHg). In contrast, postdialysis BP underestimated average ABP with wide agreement limits for both postdialysis systolic BP (33.1 to -36.3 mmHg) and diastolic BP (19.3 to -23.9 mmHg). Dialysis unit BP measurements are imprecise estimates of ABP. Better methods are needed for the assessment of BP in hemodialysis patients for clinical decision making.