Based on a novel lead compound 4-methylpiperazine-1-carbodithioic acid 3-cyano-3,3-diphenylpropyl ester 1, the systematic structural modification was carried out. All the synthesized compounds were evaluated for their in-vitro anticancer activities on four to six different cell lines at three different concentrations. Most of the tested compounds could selectively inhibit the growth of HL-60 and Bel-7402 cell lines at a medium concentration. Four compounds (3f, 3g, 3n, and 5) were selected for the IC(50) test, and the results revealed that three compounds (3g, 3n, and 5) showed almost the same or a slightly weaker activity than compound 1 against HL-60, and three compounds (3f, 3g, and 3n) showed >2-fold higher potency than compound 1 against Bel-7402. The in-vivo efficacy of 3n · HCl was evaluated with transplanted hepatocyte carcinoma 22 as an in-vivo test model. It was found that 3n · HCl could inhibit significantly the growth of tumor, and that this effect was dose-dependent. Meanwhile, the compound 3n · HCl showed low toxicity compared with compound 1 · HCl as evidenced by the little body-weight loss. These results confirmed that compound 3n · HCl is more potent than the lead compound 1 · HCl. Preliminary structure-activity relationships indicated that: a) Both nitrile group and the cyclic amine containing at least two nitrogens were indispensable moieties to keep the activity; b) substitution of the piperazine ring is unfavorable for the improvement of activity; c) the suitable linker joining the piperazinyl dithiocarboxyl and diphenylacetonitril group should be ethylene; d) a non-coplanar arrangement of the two benzene rings appears to be essential for activity.