Shortened leukocyte telomere length (LTL) is a novel biomarker for age and age-related diseases. Several epidemiological studies have examined the association between telomere length in surrogate tissues (for example, blood cells) and hypertension, and meanwhile the majority of studies reported an association some individual studies do not. We carried out a systematic review and meta-analysis to address the hypothesis that, in humans, telomere length is related with hypertension. Searches were conducted in Pubmed by September 2015 and reference lists of retrieved citations were hand searched. Eligible studies measured telomeres for both hypertensive and normotensive subjects. No restrictions were placed on sample size, publication type, age or gender. We calculated summary estimates using fixed and random effects meta-analysis. Publication bias and heterogeneity among studies were further tested. Meta-analyses from 3097 participants (1415 patients with hypertension and 1682 control subjects) showed a significant standardized mean difference between LTL in hypertensive patients and controls, either in the fixed (P < 5 × 10-6) or the random model (P < 0.005). Heterogeneity among studies was substantial (Q-statistic P-value < 0.001, I2 97.73%). Sensitivity analysis indicated that no single study changed the standardized mean difference qualitatively (0.022 > random model P-value > 0.002). Egger's test for asymmetry of effect sizes (intercept ± s.e. = − 7.278 ± 3.574; P = 0.072) did not show evidence for strong study publication bias. Leukocyte telomeres may be shorter in hypertensive than in normotensive individuals. Larger studies controlling for confounder effects are needed to confirm these findings and further explore sources of heterogeneity.