Stylistic responding is usually seen as a nuisance by researchers working with questionnaire data due to its contaminating effects on the measurement of substantiative constructs. We demonstrate that stylistic responding may be useful to improve the data quality in surveys by allowing for an identification of deviant interviewer behavior - data fabrication - in survey fieldwork. Stylistic responding in N = 710 genuine and corresponding falsified interviews was compared. Genuine survey data was collected in paper-assisted personal interviews. Corresponding falsified data were obtained by instructing falsifiers to fabricate data based on person descriptions of genuine survey respondents. Acquiescent and midpoint responding, response range, and self-enhancement emerged as useful predictors of falsification. These indicators might now be used to develop and refine multivariate statistical methods for the ex-post identification of cheating interviewers in survey fieldwork.