Background This study investigates the impact of left ventricular assist device (LVAD) inflow cannula angulation, pump positional change over time, and the incidence of thrombotic pump dysfunction in a large cohort of HeartMate II (HM2) patients. Methods Patients (n = 326) who received primary HM2 LVAD between January 2008 and December 2013 at a single institution were retrospectively reviewed. Patients who underwent pump exchange (PE) for pump dysfunction, patients who had pump dysfunction (PD) but did not require pump exchange, and patients with normal LVAD pump dysfunction (NL) were compared. Pump positional change and angle of the inflow cannula with respect to the angle between inflow cannula and the LVAD body (IL angle) were measured from routine chest radiograph at postoperation, before discharge, and follow-up. Pump positional change was assessed based on pump positional differences between discharge and follow-up. Patients were also grouped by IL acute angulation (less than 65 degrees) and pump positional change. Results There were 21, 15, and 290 patients in the PE group, PD group, and NL group, respectively. There were significant differences in IL angle between PE and NL at all timepoints: postoperation (PD 63.6 ± 12.5, NL 70.6 ± 12.3; p = 0.018), before discharge (PD 64.4 ± 12.8, NL 69.5 ± 10.5; p = 0.039), and follow-up (PD 62.6 ± 14.2, NL 67.9 ± 11.2; p = 0.002). However, the IL angle was insignificant between PE and PD groups and between PD and NL groups. Sixty-seven percent of the PE group had pump positional change as opposed to 36% of the NL group (p = 0.019). Eighty-four of 101 patients with pump positional change and 75 of 91 patients with acute angulation at postoperation did not have pump dysfunction. Conclusions Pump positional change may contribute to LVAD dysfunction or failure, but it does not entirely account for observed pump dysfunction or failure.