The validity of a black and white test box to measure changes in mouse exploratory behaviour relevant to assessment of anxiety was investigated by variation of the illumination within the test box, the use of different strains of mice, holding conditions and drug treatments. The suppression of exploratory activity in the white section caused by bright illumination was antagonised by anxiolytic agents from the benzodiazepine series, buspirone, 5-HT3 receptor antagonists, alcohol, nicotine, morphine and SCH23390. The anxiogenic agent FG7142 exacerbated the behavioural suppression. Black C57/BL/6, brown DBA2 and albino BKW mice were sensitive to the effects of drug treatments, whereas albino Tuck mice were less responsive. It is concluded that the characteristic change in mouse exploratory behaviour caused by anxiolytic agents is to preferentially increase exploratory behaviour in the white aversive section of the black and white test box. It is most consistently shown by (a) an increased time spent in the white section with proportional increases in (b) rearings and (c) ambulation and (d) a delay in the initial transition from the white to the black section.