Current modalities for predicting prostate cancer outcome are largely based on clinical features of tumors, whereas molecular markers that distinguish indolent versus aggressive tumors are lacking. Using an integrative approach that combines gene expression profiling of a mouse model with expression validation using human prostate cancer tissue microarrays, we find that the AP-1 transcription factors c-Fos and c-Jun are markers of advanced prostate cancer and in particular that c-Jun is associated with disease recurrence. Our findings provide a striking example of how accurate mouse models can inform on molecular processes involved in the progression and recurrence of human cancer. For details, see the article by Ouyang and colleagues on page 2132 of this issue.