Background In 2011, studies suggested that complications and cancer rates associated with bone morphogenetic protein (BMP) were greater than previously reported. However, later studies reported complication rates similar to prior literature and no increased cancer rate. We evaluated the pattern of clinical utilization of BMP in posteriorly based lumbar fusion by comparing two periods: 2002-2004 and 2017-2019. Methods Patients who received BMP from 2002-2004 (Early) and 2017-2019 (Late) from a single multi-surgeon institution who had a lumbar fusion were identified. One hundred patients from each cohort were randomly selected. Mean total BMP used at each level and the proportion of BMP placed in the interbody space versus posterolateral gutters were evaluated. Results In the transforaminal lumbar intebody fusion (TLIF) cohort, the total BMP dose in the Late group (6.15 mg) was nearly half of that used in the Early group (12.04 mg, p<0.000). The amount of BMP used in the posterolateral gutters remained similar (Early: 4.01 mg vs Late: 3.38 mg, p=0.222). The amount of BMP used in the interbody space was less in the Late group (2.76 mg) compared to the Early group (8.03 mg, p<0.000). In the posterior spinal fusion (PSF) cohort, the total BMP dose remained similar between the Early (11.96 mg) and the Late groups (10.82 mg, p=0.007). Conclusion Change in the use of BMP in TLIF cases was driven by the complications reported in the literature with no change in outcome. A similar impetus was not seen for PSF.