Ependymomas and tuberculomas are common paediatric posterior fossa lesions in developing countries and may be confused with each other due to some overlapping features. The distinction is important as the treatment for each is entirely different. We have described three young children in whom clinical–radiological features and the lipid peak seen on MR spectroscopy suggested the diagnosis of posterior fossa tuberculomas. All of them were started on empirical anti-tuberculosis therapy (ATT). Increase in size/no response to anti-tuberculosis therapy was disregarded as a paradoxical response. Finally, with clinical–radiological signs worsening, surgical excision was undertaken. Histopathology revealed ependymoma in all three children. This report highlights the overdiagnosis of tuberculosis in endemic areas due to biased clinical approach compounded by false positive investigations. Tissue diagnosis, though difficult by minimally invasive methods, should be sought before initiating ATT. The attempts to establish a tissue diagnosis should be continued even after starting empirical therapy, rather than waiting for the response.