To compare planned and achieved device position in patients undergoing left atrial appendage occlusion (LAAO). It is unclear how devices used for LAAO position themselves compared to what is planned. All patients undergoing LAAO at our institution had pre- and post-procedural multi detector-row computed tomography (MDCT) at 3 months (N = 52). Using dedicated software, both datasets were fused to superimpose the left atria in all planes. The effective device position was traced on the post-procedural MDCT and then imported in the pre-procedural dataset to allow comparisons. Planned and effective landing zones were compared with respect to size, location and orientation. The device’s final position was in a significantly larger landing zone than planned (452 ± 174 vs. 351 ± 112 mm² for effective and planned landing zones, respectively, paired t-test: p < 0.0001), resulting in significantly less-than-intended area oversizing (41 ± 31 vs. 12 ± 28%, p < 0.0001). In terms of device orientation, there was a difference of 19.7° between the planned and effective landing zones (p < 0.0001). The Amplatzer device had a shallower-than-planned position in 70% of cases, whereas the Watchman device had a deeper-than-planned position in 75% of cases (p = 0.04). Incomplete occlusion was found in 17 patients (33%). In a multivariable model, oversizing at the effective landing zone was the only MDCT independent predictor of incomplete occlusion (OR: 0.96 per 1% increment, 95% CI 0.95–0.98, p = 0.009). MDCT fusion showed that LAAO device position and orientation are different than planned, and this is associated with incomplete occlusion of the LAA.