Normal scapular rest position is neither adequately described nor agreed upon by authorities. The purpose of this study was to devise and describe a reliable and valid test (the Lennie Test) to measure scapular position, including normal scapular rest position. Fifteen healthy females (age 19-21 years) participated in the study. Horizontal and vertical scapular position in the frontal plane was quantified by three different investigators using a scoliometer and caliper. Same-day radiographs were used to validate scapular position surface measurements. The medial borders of the scapulae were found to be parallel to the thoracic midline. The scapulae were on average 17.19 +/- 1.85 cm apart (at the level of the root of the scapulae) with the dominant arm scapula being on average 0.49 +/- 0.74 cm lower than the nondominant scapula. This difference in height between scapulae was not statistically significant (p > .01). Correlation coefficients between skin surface and radiograph measurements of scapular position ranged from .43 to .82. Intertester intraclass correlation coefficients for surface measurements of scapular position ranged from .64 to .86. The Lennie Test was found to have moderate to high intertester reliability and to provide an accurate measurement of the anatomical location of the scapulae based on X-ray verification. Surface landmark measurements for scapular position were on average within 0.56 cm and within 1.7 degrees of the measurements made from X-rays for linear and angular position, respectively. We propose the use of the Lennie Test in populations, healthy or otherwise, where scapular position needs to be objectively measured.