Abstract Background and Aims Drug-induced acute interstitial nephritis (DI-AIN) represents a common cause of acute kidney injury. Early withdrawal of the culprit drug along with corticosteroid therapy (CS) remains the mainstay of treatment. Despite this, up to 50% of patients may not fully recover kidney function after DI-AIN. The aims of this study were to develop and externally validate a predictive nomogram to assess the probability of complete recovery (CR) of kidney function at 6 months after treatment. Method Multicenter, retrospective, observational study in 13 nephrology departments belonging to the GLOSEN group. Patients with biopsy proven DI-AIN treated with CS between 1996-2023 were included. Other potential causes of AIN such as infections or systemic diseases were carefully ruled out. Dataset was randomly divided into training group (n = 164) and validation group (n = 60). The least absolute shrinkage and selection operator (LASSO) regression was used to screen the main predictors of CR (serum creatinine value <25% of the last value before DI-AIN), and used to build the nomogram. Accuracy of the nomogram was assessed by discrimination and risk calibration, in the training and validation sets. Results The study group comprised 224 patients with DI-AIN, of whom: 51 (31%) in the training group and 19 (32%) in the validation set, achieved CR at 6 months after treatment with CS. Median age was 70 (IQR 57–76), and 115 (51%) were males. The clinical characteristics were well balanced between the training and validation data sets. The selected variables by LASSO were age (under/above 65 years), gender, degree of interstitial fibrosis and time to CS initiation (under/above 7 days). Based on multivariable logistic regression model, a nomogram was developed for assessing the probability of CR (Fig. 1). The area under the curve (AUC) of the nomogram was 0.809 (95% CI 0.721–0.880) indicating a good discriminative power. Bootstrap self- sampling was performed 1000 times for validation of the nomogram model. The AUC was 0.837 (95 % CI: 0.705–0.931), indicating good predictive stability. Calibration plot revealed that the predicted outcomes aligned well with the observations. Decision curve analysis (DCA), and clinical impact curve analysis (CICA) suggested that the model had clinical benefit. Conclusion We constructed and validated a practical nomogram with good discrimination and calibration, to predict the probability of CR of kidney function at 6 months in DI-AIN treated with corticosteroids. This clinical tool can provide clinicians with an estimate of kidney prognosis and can potentially assist them in titrating the intensity and duration of corticosteroid therapy.