Hyperinflation in patients with severe COPD is associated with an increased anteroposterior (AP) rib cage diameter. We sought to determine whether bilateral lung volume reduction surgery (LVRS) affects bony thorax configuration.Prospective of clinical data collection before and after LVRS.Tertiary-care university medical center.We measured multiple AP and transverse thoracic diameters, by using plain chest roentgenograms (CXRs) in 25 patients (11 men, 14 women), and thoracic CT scans in 14 patients (7 men, 7 women), preoperatively and 3 months postoperatively. A subgroup of 7 patients (reference data) also had CXR thoracic diameter measurements made, using films obtained previously within a year of their presurgical evaluation. Another subgroup of 10 patients had CT scan measurements also made 12 months postoperatively.CXR dimensions were taken at the level of the manubrium sterni (M) and thoracic T7 and T11 levels. CT dimensions were taken at T4, T6, T8, and T10 levels. At each level, left (L), midsagittal (C), and right (R) AP and maximal transverse diameters were measured. The sum of the three AP diameters (Total) was used for calculations. Patients also underwent tests such as spirometry, lung volumes, diffusing capacity of the lung for carbon monoxide, 6-min walk distance (6MWD), and transdiaphragmatic pressures during maximum static inspiratory efforts (Pdimax sniff) measured before and 3 months after LVRS. Patients were (mean ± SD) 58 ± 8 years old, with severe COPD and hyperinflation (FEV1, 0.68 ± 0.23 L; FVC, 2.56 ± 7.3 L; and total lung capacity[TLC], 143 ± 22% predicted). After LVRS, AP diameters were reduced at thoracic level T7 (from 24.2 ± 2.0 cm to 23.3 ± 2.2 cm, p = 0.0002), and transverse diameters were reduced at T7 (from 26.8 ± 1.9 cm to 26.4 ± 1.7 cm, p = 0.001) and T11 (from 29.9 ± 2.2 cm to 29.5 ± 2.2 cm, p = 0.03), as measured using the CXR. In contrast, thoracic diameters were similar in subjects with CXRs before LVRS and within 1 year before evaluation. CT-measured AP diameters were significantly reduced 3 months after LVRS at T6, (from 48.8 ± 6.0 cm to 46.7 ± 5.4 cm, p = 0.02), T8 (from 54.2 ± 7.0 cm to 52.3 ± 6.5 cm, p = 0.004), and T10 (from 53.8 ± 7.5 cm to 51.2 ± 8.0 cm, p = 0.001), but not at T4. These AP diameter reductions directly correlated with the postoperative reductions in TLC and residual volume, and also with the increases in Pdimax sniff and 6MWD after LVRS. The reduction in AP diameters at thoracic levels T8 and T10 seen 3 months after LVRS remained stable at 12-month follow-up, whereas those measured at T6 lost statistical significance. CT-measured transverse diameters were unchanged at all levels after LVRS.We conclude that LVRS decreases mid-to-lower AP rib cage diameter as assessed by CXR and thoracic CT. Although transverse diameters were reduced on CXR, the magnitude was small and was not confirmed with CT. After LVRS, AP diameter reductions are most likely the result of reduction in lung volume, and they are associated with improvements in diaphragm strength and exercise endurance.