The relationship between Fassier-Duval (FD) rod placement and rod failure rates has not previously been quantified. Retrospective review was conducted on patients with osteogenesis imperfecta treated with FD rods between 2005 and 2017. Age at first surgery, sex, Sillence type of osteogenesis imperfecta, bisphosphonate treatment, location of rod (side of body and specific bone), and dates of surgeries, radiographs, and rod failures were collected. C-arm images determined rod fixation within the distal epiphysis at the time of surgery. C-arm variables included rod deviation (percent deviation from the midline of the distal epiphysis) and anatomical direction of deviation (anterior/posterior and medial/lateral). X-ray images were examined for rod failure, which was defined as bending, pulling out of the physis, protrusion out of the bone, and/or failure to telescope. Cox proportional hazards regression models were used to compare failure rates with location of placement within the distal epiphysis allowing for clustering of the data by side (left or right) and bone (femur or tibia). The cohort was 13 patients (11 female individuals and 2 male individuals) with a total of 66 rods and 75 surgeries. Mean time from the first surgery to the last follow-up visit was 8.9 years (SD=5 y). There was a 7% increase in hazard of failure per 1-mm increase in antero-posterior (AP) deviation [hazard ratio (HR), 1.07; 95% confidence interval (CI), 1.01-1.14; P=0.029)]. Similarly, there was a 9% increase in hazard of failure for every 1-mm increase in lateral deviation (HR, 1.09; 95% CI, 1.01-1.18; P=0.019). A 12% increase in hazard of failure per 10% increase in deviation from the midline for both AP and lateral radiograph views was also found, although this was only statistically significant for lateral deviation on the AP radiograph view (HR, 1.12; 95% CI, 1.01-1.25; P=0.030). FD rod placement within the distal epiphysis has significant impact on increasing rod survival. Level III-therapeutic study.