Interest in 2'-deoxyribonucleosides and arabinofuranosyl nucleosides as potential anticancer chemotherapeutic agents was stimulated by the antileukemic activity of 1-β-D-arabinofuranosylcytosine (ara-C), though ara-C has drawbacks including a short plasma half-life due to rapid deamination and ineffectiveness against solid tumors. Previous studies on 2'-branched-chain nucleosides such as (2'S)-2'-deoxy-2'-C-methylcytidine (If, SMDC) and 2'-deoxy-2'-methylidenecytidine (2, DMDC) showed potent antineoplastic activity, with their 5'-triphosphates inhibiting DNA polymerases and acting as chain terminators. To design a nucleoside that induces DNA strand scission, we developed 2'-C-cyano-2'-deoxy-1-β-D-arabinofuranosylcytosine (CNDAC), hypothesizing that the 2'-C-cyano group would enable β-elimination (analogous to β-cyanoethyl phosphate deblocking) upon incorporation into DNA, causing strand breakage. CNDAC was synthesized using a radical deoxygenation methodology for 2'-branched nucleosides. In vitro, CNDAC inhibited mouse leukemic L1210 cell growth with an IC50 of 0.21 μg/mL (comparable to DMDC and SMDC) and showed potent cytotoxicity against 14 of 16 human tumor cell lines (IC50 0.04-6.8 μg/mL), including all tested stomach adenocarcinomas (high incidence in Japan), while ara-C was effective in only 6 lines. In vivo, CNDAC significantly prolonged survival of P388 leukemia-bearing mice: at 100 mg/kg i.p. on days 1 and 5, the median survival ratio (treated vs. control) was 183% (vs. 163% for ara-C); at 20 mg/kg i.p. on days 1-10, 5 of 6 mice survived over 60 days. CNDAC is a promising agent for further evaluation, and investigations into its mechanism (whether DNA strand breakage is responsible) are ongoing, which could represent a novel approach to anticancer chemotherapy if confirmed.