Objective To evaluate the predictive value of four-dimensional automatic left atrial quantification (4D-LAQ) on cardiovascular complications in patients with diabetic nephropathy (DN). Methods 120 patients with DN in the hospital were selected from March 2022 to December 2023. According to whether cardiovascular complications occurred, they were divided into cardiovascular disease group (n=51) and non-cardiovascular disease group (n=69). Another 50 healthy subjects with physical examination during the same period were included in control group. The general data and conventional echocardiographic parameters [left ventricular end-diastolic diameter (LVEDD), left atrial anteroposterior diameter (LAD), left ventricular ejection fraction (LVEF), left ventricular end-systolic diameter (LVESD), early diastolic peak velocity of mitral annulus (E/A)] and 4D-LAQ parameters [maximum left atrial volume (LAVmax), minimum left atrial volume (LAVmin), maximum left atrial volume index (LAVImax), left atrial pre-systolic volume (LAVpreA), left atrial total emptying fraction (LAEF), left atrial systolic longitudinal strain (LASct), left atrial conduit longitudinal strain (LAScd) and left atrial reserve longitudinal strain (LASr)] were compared among the groups. The correlation between left atrial strain parameters and volume parameters and the diagnostic efficiency of each parameter on cardiovascular complications in DN were analyzed. Results There were significant differences in LVEDD, LAD, LVEF, LVESD and E/A in cardiovascular disease group and non-cardiovascular disease group compared with those in control group (P<0.05). Significant differences were shown in LVEDD, LAD, LVEF and LVESD between cardiovascular disease group and non-cardiovascular disease group (P<0.05). The LAVmax, LAVmin, LAVImax, LAVpreA, LAEF, LASct, LAScd, LASr, 4D-GRPS and 4D-GLPS in cardiovascular disease group and non-cardiovascular disease group were obviously different from control group (P<0.05), and the above parameters between cardiovascular disease group and non-cardiovascular disease group were significantly different (P<0.05). Pearson correlation analysis showed that LVEDD and LAD in cardiovascular disease group were positively correlated with LAVmax, LAVmin, LAVImax, LAVpreA, LAScd and 4D-GLPS(P<0.05), and were negatively correlated with LAEF, LASct, LASr and 4D-GRPS(P<0.05). LVESD and LVEF were negatively correlated with LAVmax, LAVmin, LAVImax, LAVpreA, LAScd and 4D-GLPS(P<0.05), and were positively correlated with LAEF, LASct, LASr and 4D-GRPS(P<0.05). ROC curve revealed that the AUCs of 4D-LAQ parameters LAVmax, LAVmin, LAVImax, LAVpreA, LAEF, LASct, LAScd, LASr, 4D-GRPS和4D-GLPS were 0.764, 0.649, 0.778, 0.649, 0.666, 0.895, 0.776, 0.697, 0.842 and 0.865 respectively. The AUCs of conventional ultrasound parameters LVEDD, LAD, LVESD and LVEF were 0.718, 0.693, 0.610 and 0.497 respectively. Conclusion Four-dimensional automatic left atrial quantification can effectively evaluate the changes of left atrial volume and function in patients with DN, and has a high predictive value on evaluating cardiovascular complications in patients with DN.