Telomerase activity (TA) is the most recently recognized prognostic factor in neuroblastoma, and its outstanding predictive power was documented by several studies. However, TA measurements require fresh tumor tissue that is not always available in daily clinical practice. We previously described a reverse transcriptase-polymerase chain reaction assay that we used to investigate the possible prognostic relevance of the telomerase catalytic subunit, hTERT, at the mRNA level. Because hTERT mRNA undergoes alternative splicing as a regulatory mechanism of TA, we discriminated between truncated and full-length hTERT transcripts. In a retrospective study on 124 neuroblastomas, 56 (45.2%) tumors showed spliced hTERT transcripts, whereas 30 (24.2%) contained full-length hTERT transcripts. The presence of both spliced and full-length hTERT transcripts was significantly associated with MYCN amplification. hTERT in general showed no correlation to other prognostic factors, ie, International Neuroblastoma Staging System stage, International Neuroblastoma Pathology classification grade, or age at diagnosis, whereas the presence of full-length transcripts was significantly associated with higher stages. The presence of any hTERT transcripts carried no significant prognostic information, yet full-length hTERT transcripts were highly predictive of poor outcome (P < 0.0001). In a multivariate analysis, full-length hTERT transcripts and International Neuroblastoma Pathology classification grade emerged as the sole independent predictors of event-free survival, with relative risks of 10.0 and 3.9, respectively. The strong statistical correlation of full-length hTERT transcripts with clinical outcome in neuroblastoma suggests that the reverse transcriptase-polymerase chain reaction analysis of hTERT transcripts may be equatable to TA measurements. Because this assay is well suited for archival material, it could become a useful adjunct in evaluating the prognosis of individual neuroblastoma cases.