The Implicit Relational Assessment Procedure (IRAP) was designed to detect relational responding that cannot easily be accessed via traditional survey methods. The IRAP requires participants to meet speed and accuracy criteria during practice trials before proceeding to test trials, which has resulted in an attrition rate of approximately 20%, on average, in the existing research. Variables affecting the attrition rate have not been systematically investigated. I examined the influence of incentives (in this case a $20 voucher contingent on meeting performance criteria) on attrition rate and other IRAP performance measures. In addition, I examined whether the IRAP would reveal an implicit anti-fat bias in 82 university students. I found significant differences in the performance of the incentive group compared to the control group in their response accuracy and measurement of their implicit bias. The results indicated higher levels of bias compared with those from previous research studies, particularly in the incentive group. I did not find statistically significant differences in the attrition rate but found a low attrition rate in both groups. This study reveals the utility of incentives for improving performance on the IRAP, a procedure that demands accurate responses under time pressure for assessing spontaneous relational responding.