We randomized half of the sample from two southern California medical centers to a post-paid incentive (n = 1,795) and half to no incentive (n = 1,797) for completing a web-based survey about their experiences with health care. Respondents in the incentive group were given the choice between a $5 cash or Target® e-certificate. The characteristics of respondents in the incentive and control groups was similar on age, education, length of membership in the plan, number of emails sent and visits to the primary care doctor in the 12 months prior to sampling, and their global rating of the doctor; the incentive group had more Asians (8% vs. 5%, χ 2 (1 df) = 7.92, p = 0.005) and fewer Blacks/African Americans (2% vs. 4%, χ 2 (1 df) = 11.0, p =0.001) than the no incentive group. Those randomized to the incentive were significantly more likely to respond to the survey than those in the control group (57% vs. 50%, t (df = 3590) = 4.06, p <0.0001). Item non response rates were similar for those in the incentive and the control groups. Those randomized to the incentive condition who completed the survey were more likely to prefer a cash incentive over the e-certificate (69% of the incentives delivered to web respondents were in the form of cash). The unit cost per incentive was $8.32 for cash and $7.49 for the e-certificate. The results of this experiment indicate that a post-paid incentive can significantly increase the response rate to a web-based survey.