Online testing is now becoming more prevalent internationally, across different educational levels. However, little is known about the impact of this mode of testing on male and female students' anxiety levels and test performance. Conducted in an Indonesian sociocultural context which is under-represented in the international literature, this paper presents results from a study examining the effect of two different modes of testing (online and paper-and-pencil tests) on test takers' anxiety levels and test scores. A total of 243 (99 males, 144 females) junior high school students were recruited. These students were assigned to either paper-and-pencil or online testing. The PHCC test anxiety questionnaire was administered immediately after the test. A 2 x 2 factorial ANOVA revealed that neither gender, mode of test administration nor the interaction between gender and mode of testing impacted on students' anxiety levels. The main effect of gender and the interaction effect between gender and mode of testing on test scores were significant. Students who experienced high anxiety levels performed significantly poorer than those who had medium or low anxiety levels.