Most human somatic cells do not divide indefinitely but enter a terminal growth arrest termed replicative senescence. Replicatively senescent cells are generally believed to arrest in G1 or G0 stage of the cell cycle. While doing cell cycle analysis on three different lines of normal human fibroblasts we observed that 36-60% of the replicatively senescent cells had 4N DNA content. Only up to 5% of senescent cells had more than one nucleus ruling out the possibility that the 4N cell population were G1-arrested bi-nucleated cells. Furthermore, it is unlikely that the 4N cells are tetraploids, because actively dividing pre-senescent cultures lacked the 8N tetraploid G2 population. Collectively these results suggest that the 4N population consists of G2 arrested cells. The notion that a large fraction of senescent cell population is arrested in G2 is important for understanding the biology of replicative senescence.