The objectives of this study were to determine both the intra-rater reliability and the smallest real difference necessary to detect meaningful clinical changes over an 8-12 week period for three clinical measures of posterior shoulder flexibility. Posterior shoulder tightness has been associated with abnormal humeral head translations that narrow the subacromial space and contribute to impingement. Posterior shoulder stretching to normalize joint kinematics is often used when treating individuals with subacromial impingement. Currently there is no measurement of posterior shoulder flexibility that has been demonstrated to be reliable or responsive in patients over a treatment interval. Construction workers with overhead work exposure were recruited to an impingement group (n=37), and an asymptomatic group (n=22). Three flexibility measurements were taken on each subject at baseline and 8-12 weeks later by the same physical therapist. Intra-rater reliability was determined using intra-class correlation coefficients (ICCs) and standard errors of measurement (SEM). The smallest real difference value (SRD) determined the smallest detectable intervention-related measurement change over time. ICCs for the three measurements ranged from poor (0.40) to good (0.79). The SEM and SRD values reflected high test-retest variability in all three measurements. None of the three measures were proven to be highly stable indicators of posterior shoulder flexibility over this time period, nor were they able to detect small changes over this interval due to the high trial to trial variance in a non-intervention sample. Caution should be used when interpreting these clinical measurements over prolonged time periods.