Objective  To investigate the expression and clinical significance of platelet parameters and neutrophil-to-lymphocyte ratio (NLR) in diabetic nephropathy (DN) patients under peritoneal dialysis (PD). Methods  A retrospective analysis was conducted on 83 DN patients who underwent PD treatment in our hospital from February 2020 to June 2022. The patients' general information and laboratory indicators after admission were collected, and the patients were divided into survival group and death groups according to their survival outcomes within 2 years. The receiver operating curve (ROC) was used to analyze the predictive value of platelet-related parameters and NLR on the prognosis of PD in patients with DN, and multivariate logistic regression was used to analyze the influencing factors of patient death.  Results  Compared with the death group, the mean platelet volume (MPV), platelet distribution width (PDW), platelet count (PLT), NLR, and blood urea nitrogen (BUN) and serum creatinine (Scr) were lower in the survival group (t=6.071, 5.093, 3.730, 3.604, 19.166, 23.365, P<0.001, <0.001, <0.001, <0.001, <0.001, <0.001). Multivariate logistic regression analysis pointed out that high levels of PLT (OR=1.879, 95% CI:1.355～2.607, P＜0.001), MPV (OR=2.063, 95% CI:1.504～2.828, P＜0.001), PDW (OR=1.998, 95% CI:1.449～2.755, P＜0.001), NLR (OR=1.828, 95% CI:1.282～2.606, P=0.001), Scr (OR=2.924, 95% CI:1.712～4.993, P＜0.001), BUN (OR=2.746, 95% CI:1.653～4.561, P＜0.001) are risk factors for death after PD in DN patients. ROC curve analysis suggested PLT (AUC=0.714, P=0.007), MPV (AUC=0.788, P＜0.001), PDW (AUC=0.750, P=0.002), and NLR (AUC=0.741, P=0.002) as predictors of the outcome of DN patients. The sensitivity and specificity (AUC=0.838, P＜0.001) were 89.4% and 88.2%, respectively.  Conclusion  MPV, PDW, PLT, NLR, Scr, and BUN levels were the contributing factors for death after PD in DN patients. The expression levels of MPV, PDW, PLT and NLR can effectively predict the survival of DN patients treated with PD, and the combined prediction efficacy is higher.