The present invention relates to a scoring system for the prediction of cancer recurrence. More particularly, the present invention concerns with the selection of genes and/or proteins, and generation of formulae with the selected genes and/or proteins for the prediction of cancer recurrence by measuring the expression of genes and/or proteins of human tumor tissues, and comparing their patterns with those of the gene and/or protein expression of human primary tumors from patients who have cancer recurrence and those who do not have cancer recurrence. The present invention also relates to a kit for performing the method of the present invention comprising DNA chip, oligonucleotide chip, protein chip, peptides, antibodies, probes and primers that are necessary for effecting DNA microarrays, oligonucleotide microarrays, protein arrays, northern blotting, in situ hybridization, RNase protection assays, western blotting, ELISA assays, reverse transcription polymerase-chain reaction (hereinafter referred to as RT-PCR) to examine the expression of at least 2 or more of genes and/or proteins, preferably 4 or more of genes and/or proteins, more preferably 6 or more of genes and/or proteins, and most preferably 12 or more of genes and/or proteins, that are indicative of cancer recurrence.