The relevance of high-dose chemotherapy followed by auto-SCT in CLL remains to be defined. The aim of the prospective, randomized, GOELAMS LLC 98 trial was to compare two strategies in previously untreated CLL patients aged <60 years. Conventional chemotherapy (Arm A) consisted of six monthly courses of CHOP followed by six CHOP courses in every 3 months in those achieving a complete or PR. Arm A was compared with high-dose therapy with auto-SCT (Arm B), used as consolidation after three CHOP courses in case of CR or very good PR. A total of 86 patients were enrolled, of which 39 and 43 patients were evaluable in arm A and arm B, respectively. The primary endpoint was PFS. On an intent-to-treat basis and with a median follow-up time of 77.1 (range 1-135.5) months, the median PFS was 22 months in Arm A and 53 months in Arm B (P<0.0001). Median survival time was 104.7 months in arm A and 107.4 months in arm B. This trial demonstrates that frontline high-dose therapy with auto-SCT prolongs PFS but does not translate into a survival advantage in advanced CLL patients in the pre-rituximab era.