Purpose: Certain amounts of clavicle fracture displacement have been associated with poor functional outcome, making the ability to measure this displacement an important diagnostic goal. There is variability, however, in radiographic technique on shoulder images. Additionally, it has been shown that a subset of clavicle fractures progressively displace in the peri-injury time period. The purpose of this study is to (1) describe the incidence of progressive displacement greater than 1 cm in the peri-injury time frame, (2) compare different methods of measuring clavicle displacement using interobserver reliability, and (3) determine whether supine versus upright position during radiography changes measured displacement. Methods: This was a prospective observational study in which 60 consecutive patients diagnosed with a displaced midshaft clavicle fracture had a specific radiographic protocol. All patients presenting with a midshaft clavicle fracture within 7 days of injury met inclusion. In addition, inclusion required at least one follow-up with protocol-defined radiographs. The protocol included 3 views: (1) standing 15° cephalic tilted AP, (2) supine 15° cephalic tilted AP clavicle, and (3) supine 15° cephalic tilted AP panoramic shoulder girdle view. Vertical displacement and clavicle shortening was independently measured by 2 trained examiners at all time points to detect changes resulting from patient positioning and time from injury. To analyze whether time from injury affected the difference between upright and supine displacement measurements, radiographs were categorized into 3 groups: (1) images taken on days 0-6 postinjury (n = 66), (2) days 7-21 postinjury (n = 46), and (3) days 22 or greater postinjury (n = 29). Results: 60 patients had initial films at a mean of 1 day after injury with the second set of films done at a mean of 12 days after the initial radiographs. 30 patients had a third set of radiographs done at a mean of 41 days after the initial radiographs. Between the initial and the final follow-up radiograph, 6 patients (10%) had progressive medialization greater than 1 cm and 13 patients had progressive translation (22%) greater than 1 cm. The interobserver reliability as determined by the concordance correlation coefficient was nearly perfect for the measurement of medialization (0.929), translation (0.982), and length of injured clavicle (0.996). Analysis showed that upright radiographs had 11%, 27%, and then 10% greater medialization compared to supine in the respective time categories (0-6 days, 7-21 days, 22+ days). Upright images also showed 16%, 19%, and then 9% greater translation than supine in the respective time categories. Statistical analysis comparing medialization and translation from upright to supine at all time points revealed significant differences (P <0.05). These data would suggest that the difference between upright and supine displacement measurements diminished over time as would be expected due to consolidation of the fracture. Conclusion: Our data show that close follow-up of nonoperatively treated clavicle fractures is warranted due to the risk of progressive displacement. The data also show that upright gravity views highlight displacement both for medialization and translation of the fracture. We therefore recommend consideration of a change in radiographic protocol that mandates upright films when possible, since measurements of displacement in this position may affect surgical decision making.