Background: Renal abscess (RA) is rarely seen in the paediatric age group. The proposed management protocols are mainly derived from the adult series which may not be appropriate in children. Objective: In this retrospective analysis of cases with renal and perinephric abscesses, the objective is to correlate the clinical presentation, radiological findings and treatment options and also to propose a paediatric-specific practical management algorithm. Study design: This is a retrospective study of cases with renal and perinephric abscesses admitted between March 2012 and February 2020. The patients were reviewed for demographics, presentation, predisposing factors, laboratory investigations, imaging, management and outcome. Results: Analysis of 12 paediatric patients (13 renal units) with RA (median age 4 years) was done. Organisms were isolated in 8 of 12 (66.6%) patients with Gram-negative organisms being the commonest. On admission, all patients were started on empirical broad-spectrum antibiotics. Except for two patients who were critically ill with frank sepsis and had a tender renal lump, the rest of them were initially offered conservative management with intravenous antibiotics, and the response was reviewed after 48–72 hours. Of five units with abscess size of ⩽3 cm, two units (40%) responded to conservative management, while three units (60%) required intervention, and of eight units of size >3 cm, three units (37.5%) responded to conservative management and five units (62.5%) required intervention. None of the abscesses with perinephric collection (30.7%) responded to antibiotics and required intervention. Conclusion: A protocol based on the size of RA as recommended in most of the adult series may not be appropriate in the paediatric age group because of the differences in clinical presentation, predisposing factors and immune response. The clinical condition on presentation, response to antibiotic therapy and the presence of perinephric collection should be considered as an important determinant in deciding the need for intervention. Level of evidence: 4