Background: Pathological tendons are known to increase their diameter, which can be viewed using ultrasound (US) and MRI. Although systematic reviews have confirmed the reliability of US for the measurement of muscle thickness, no such reviews exist to examine the reliability of US measures of tendon size. Purpose: To systematically review the literature on the reliability of US to assess tendon dimensions including thickness and cross-sectional area in human limbs. Methods: A comprehensive review of electronic databases was performed by two reviewers using an agreed range of keywords. Clinical trials which investigated inter or intra rater reliability of real time ultrasound was included in this review. The Quality Appraisal of Reliability Studies (QAREL) checklist was used to assess risk of bias. Included studies performed ultrasound analysis in a range of clinical settings. Variations existed between studies in terms of rater training, tendons studied, and participant pain status. Inter and Intra rater reliability of tendon thickness and cross sectional area were analysed using estimates of both reliability and precision using intraclass correlation coefficients (ICC’s) and coefficients of variation (CVs), percentage coefficient of variation (CV%), percentage standard error of the mean (SEM%), limits of agreement (LOA), pearsons correlation coefficient (r2) and 95% confidence intervals (95% CI). Results: A total number of 615 were included in the review. Of this 136 were Symptomatic and 479 were asymptomatic participants, with an age range: 17.5–73 years. Assessment of the risks of bias indicated 15/16 chosen studies was of high quality. ICC values for inter and intra rater reliability indicate good to excellent reliability for all tendon sites with values ranging from (0.55–0.99). Percentage coefficent of variation (CV%) ranged from 0-16% across all tendon sites. The most commonly studied tendons achilles and supraspinatus, the average measurement error from the included studies on Achilles tendons was 9.5% %CV and 5.39% SEM%, these figures equate to errors of 0.67mm and 0.32mm respectively which is well below the observed differences between symptomatic and asymptomatic tendons. The average measurement error from the included studies on supraspinatus tendons was 11.24% %CV and 5.72% SEM%, which equate to errors of 0.77mm and 0.39mm respectivelywhich is similar to, or below, the proposed differences between symptomatic and asymptomatic tendons.LOA ranged from (0.35–3.50mm), r2 ranged from (0.43–0.92), 95% CI ranged from (0.43–0.98) for all tendons. Conclusion(s): Findings of the current review appear to indicate that real time ultrasound is a clinically reliable method inmeasuring tendondimensions.However thesefindings must be considered in light of the limitations identified in the studies with a wide array of statistical methodology and wide range of measurement protocols and assessors. Implications: The findings of this current review may have important clinical implications as regards indicating future direction for possible injury prevention strategies. An understanding of the relationship between sonographic abnormalities and the likelihood of the development of symptoms in asymptomatic individuals, may allow ultrasound to be used as a screening tool to identify athletes at risk and guide prophylactic treatments or training modifications.