Background: The cyclin-dependent kinase inhibitor p27 Kip1 functions during normal cerebellar development and has demonstrated tumor suppressor functions in mouse models of medulloblastoma. Because P27 loss is associated with increased proliferation, we assessed whether P27 absence in surgical medulloblastoma specimens correlated with response to therapy in pediatric patients enrolled in two large studies. Additionally, we examined the functional consequence of p27 Kip1 loss in the SmoA1 medulloblastoma model to distinguish whether p27 Kip1 reduces tumor initiation or slows tumor progression. Findings: Analysis of 87 well-characterized patient samples identified a threshold of P27 staining at which significant P27 loss correlated with poor patient outcome. The same criteria, applied to a second test set of tissues from 141 patients showed no difference in survival between patients with minimal P27 staining and others, suggesting that P27 levels alone are not a sufficient prognostic indicator for identifying standard-risk patients that may fail standard therapy. These findings were in contrast to prior experiments completed using a mouse medulloblastoma model. Analysis of cerebellar tumor incidence in compound mutant mice carrying the activated Smoothened (SmoA1) allele that were heterozygous or nullizygous for p27 Kip1 revealed that p27 Kip1 loss did not alter the frequency of tumor initiation. Tumors haploinsufficient or nullizygous for p27 Kip1 were, however, more invasive and displayed a higher proliferative index, suggesting p27 Kip1 loss may contribute to SmoA1 medulloblastoma progression. Conclusions: These studies revealed P27 loss affects medulloblastoma progression rather than initiation and that this putative biomarker should not be used for stratifying children with medulloblastoma to risk-based therapeutic regimens.