BACKGROUND Current guidelines are inadequate for use in predicting ITP recurrence. Therefore, our primary goal in this study was to investigate the association of platelet-to-lymphocyte ratio (PLR) at diagnosis with ITP recurrence in Chinese patients. MATERIAL AND METHODS We performed a historical cohort study and non-selectively enrolled 233 patients with newly-identified ITP from March 2013 to June 2017. The independent variable was PLR recorded at diagnosis and the dependent variable was recurrence-free survival (RFS) at 6 months. Data on the following variables were also collected for establishing a multivariate Cox regression model: demographic details, general details, and variables found to be closely related to PLR in previous studies, as well as risk factors for ITP recurrence. RESULTS During follow-up, 85 patients had an event within 6 months. At the range of 0.86-9.7 of PLR, a 1-unit increase in PLR was associated with a 13% decrease in ITP recurrence (hazard ratio: 0.87; 95% confidence interval: 0.78-0.97), whereas no association was detected at the range of 9.7-33.75 of PLR (hazard ratio: 0.99; 95% confidence interval: 0.95-1.04). An interaction test indicated that patients with HP infection (0.91 (0.86-1.97)) or diabetes history (0.86 (0.78-0.96)) showed a stronger association compared with patients without HP infection (1.01 (0.95-1.04) and those without diabetes (1.01 (0.97-1.04)). CONCLUSIONS Our findings suggest that PLR is a useful parameter to consider when hematologists attempt to assess the risk of recurrence in ITP patients receiving first-line therapy, and the nonlinearity of PLR and ITP recurrence risk must be fully considered when constructing predictive models.