To examine the influence of data anonymity on socially desirable responding, the Marlowe-Crowne Social Desirability Scale, the Perceived Stress Scale, and a physical symptoms checklist were administered to 140 Chinese undergraduates under conditions of Anonymity or Nonanonymity. The degrees of association between any two of these three measures were not significantly different between the two conditions. Moreover, the present sample of Chinese subjects did not exhibit accentuation in socially desirable responding. Despite the nonsignificant effect of anonymity on social desirability response, future stress research should not overlook the possible impact of this factor on data validity since difference between two correlation coefficients is difficult to detect, and Nonanonymity did considerably inflate the correlation between stress and symptom scores.