To establish the cutoff values for screening ENU-induced behavioral mutations, normal variations in mouse behavioral data were examined in home-cage activity (HA), open-field (OF), and passive-avoidance (PA) tests. We defined the normal range as one that included more than 95% of the normal control values. The cutoffs were defined to identify outliers yielding values that deviated from the normal by less than 5% for C57BL/6J, DBA/2J, DBF(1), and N(2) (DXDB) progenies. Cutoff values for G1-phenodeviant (DBF(1)) identification were defined based on values over +/- 3.0 SD from the mean of DBF(1) for all parameters assessed in the HA and OF tests. For the PA test, the cutoff values were defined based on whether the mice met the learning criterion during the 2nd (at a shock intensity of 0.3 mA) or the 3rd (at a shock intensity of 0.15 mA) retention test. For several parameters, the lower outliers were undetectable as the calculated cutoffs were negative values. Based on the cutoff criteria, we identified 275 behavioral phenodeviants among 2,646 G1 progeny. Of these, 64 were crossed with wild-type DBA/2J individuals, and the phenotype transmission was examined in the G2 progeny using the cutoffs defined for N(2) mice. In the G2 mice, we identified 15 novel dominant mutants exhibiting behavioral abnormalities, including hyperactivity in the HA or OF tests, hypoactivity in the OF test, and PA deficits. Genetic and detailed behavioral analysis of these ENU-induced mutants will provide novel insights into the molecular mechanisms underlying behavior.