This study was undertaken to evaluate the reproducibility (interobserver variability), and the short-term and long-term repeatability (intraobserver variability) of sonographic measurement of the maximum and mean splenic length. Ultrasound (US) measurements of the splenic length were performed in 43 subjects, ages 30 to 59 years old (mean age = 44.5 years). Each subject attended three US sessions: baseline examination and 30 min and 8 weeks after the baseline examination. In each session, US examinations of the spleen were performed and the splenic length was measured by four operators. In each examination, the splenic length was measured 3 times, and the maximum and mean values were obtained. The overall reproducibility of the measurements of maximum and mean splenic lengths were 67% and 89%, respectively. The short-term repeatability in measuring the maximum and mean splenic lengths were 87% and 94%, respectively, whereas the long-term repeatabilities were 61% and 76%, respectively. The reproducibility and short-term and long-term repeatability of the measurement of mean splenic length were higher than those of the maximum splenic length. In the mean and maximum splenic length measurement, the short-term repeatability was higher than the long-term repeatability. The results suggested that the mean splenic length has a higher reliability in sonographic measurement and should be used in routine clinical practice. Measurement errors should be considered when evaluating the changes of splenic length in serial US examinations, particularly in long-term follow-up. (E-mail: ormying@polyu.edu.hk) ; School of Optometry