A continuous marine fish cell line RSBF (i.e. Red Sea Bream Fin) was utilized to screen the cytotoxicity and genotoxicity of polyethylenimine (PEI) and nickel chloride (NiCl 2) in this study on the deleterious effects of aquatic genotoxins on fish. At the 0.01 to 1 μg/ml concentration tested, PEI had acute toxicity to the treated RSBF cells (IC 50 =1.12, 0.92, 0.88 and 0.64 μg/ml PEI for time 0 h, 24 h, 48 h and 72 h after treatment, respectively) and markedly inhibited their proliferation in a dose dependent manner. At the 0.001 to 5 μmol/L concentration tested, NiCl 2 posed no acute toxicity but significantly stimulated their growth (107%-214% of control). Random amplified polymorphic DNA (RAPD) technique was used to detect the genotoxic effects of PEI and NiCl 2 by comparing the RAPD banding patterns of the control and treated cells. RAPD analysis indicated that at the concentrations tested, PEI was more genotoxic than NiCl 2 to RSBF cells; that there was a slight dose dependent response in the genotoxic effect of PEI but not NiCl 2; and that RAPD technique might provide a sensitive, non specific genotoxic endpoint. And the potent cytotoxicity and genotoxicity of PEI on fish cells showed that we should be cautious in utilizing it as gene vector in fish gene transfer and human gene therapy. A continuous marine fish cell line RSBF (ie Red Sea Bream Fin) was utilized to screen the cytotoxicity and genotoxicity of polyethylenimine (PEI) and nickel chloride (NiCl 2) in this deleterious effects of aquatic genotoxins on fish. At the 0.01 PEI had acute toxicity to the treated RSBF cells (IC 50 = 1.12, 0.92, 0.88 and 0.64 μg / ml PEI for time 0 h, 24 h, 48 h and 72 h after treatment, respectively) and markedly inhibited their proliferation in a dose dependent manner. At the 0.001 to 5 μmol / L concentration tested, NiCl2 posed no acute toxicity but significantly stimulated their growth (107% -214% of control). Random amplified polymorphic DNA (RAPD) technique was used to detect the genotoxic effects of PEI and NiCl 2 by comparing the RAPD banding patterns of the control and treated cells. RAPD analysis indicated that at the concentrations tested, PEI was more genotoxic than NiCl 2 to RSBF cells; that there was a slight do se dependent response in the genotoxic effect of PEI but not NiCl 2; and that RAPD technique may provide a sensitive, non specific genotoxic endpoint. And the potent cytotoxicity and genotoxicity of PEI on fish cells showed that we should be cautious in utilizing it as gene vector in fish gene transfer and human gene therapy.