The Texas Parks and Wildlife Department conducted 2 harvest surveys by mail during 1976, I sampled currently licensed hunters and the other sampled hunters that were licensed I year previously. Both surveys requested information concerning hunting activity during the current season. A portion of the currently licensed hunters who did not return questionnaires were contacted by telephone to adjust for non-response bias. The survey sampling out-of-date hunters was not adjusted for non-res pone bias. The harvest and related hunting statistics estimated from regression models developed from the result of these 2 surveys are comparable to statistics estimated by sampling currently licensed hunters. This technique also provides for the early availability of harvest statistics (6 weeks after the last day of the season) and reduces the expense ofconducting the survey by eliminating the telephoning requirement to adjust for non-response.