Cage subsidence (CS) has been reported to be one of the most common complications following oblique lumbar interbody fusion (OLIF). To reduce the incidence of CS and improve intervertebral fusion rates, anterolateral fixation (AF) has been gradually proposed. However, the incidence of CS in patients with oblique lumbar interbody fusion combined with anterolateral fixation (OLIF-AF) is still controversial. Additionally, there is a lack of consensus regarding the optimal placement of screws for OLIF-AF, and the impact of screw placement on the incidence of CS has yet to be thoroughly investigated and validated. The objective of this investigation was to examine the correlation between screw placements and CS and to establish an optimized approach for implantation in OLIF-AF. A retrospective cohort study was undertaken. From October 2017 to December 2020, a total of 103 patients who received L4/5 OLIF-AF for lumbar spinal stenosis or spondylolisthesis or degenerative instability in our department were followed up for more than 12 months. Demographic and radiographic data of these patients were collected. Additionally, screw placement related parameters, including trajectory and position, were measured by anterior-posterior X-ray and axial CT. Analysis was done by chi-square, independent t-test, univariable and multivariable binary logistic regression to explore the correlation between screw placements and CS. Finally, the receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive ability of screw placement-related parameters. A total of 103 patients were included, and CS was found in 28 (27.18%) patients. Univariable analysis was firstly performed for each parameter. Next, variables with p-value of <0.05, including bone mineral density (BMD), concave morphology, and screw placement-related parameters were included in the multivariate logistic regression analysis. Significant predictor factors for subsidence were coronal plane angle (CPA) (OR 0.580 ± 0.208, 95% CI 1.187-2.684), implantation point (IP) (L4) (OR 5.732 ± 2.737, 95% CI 1.445-12.166), and IP (L5) (OR 7.160 ± 3.480, 95% CI 1.405-28.683). Furthermore, ROC curves showed that the predictive accuracy of CS was 88.1% for CPA, 77.6% for IP (L4) and 80.9% for IP (L5). We demonstrate that the trajectory of vertebral screws, including angle and position, was closely related to CS. Inserting screws parallel to each other and as close to the endplate as possible while keeping the cage inside the range of the superior and inferior screws are an optimal implantation strategy for OLIF-AF.