We consider estimation of the population total in double sampling for regression in the presence of nonresponse. Two estimators were examined, one allowing for nonresponse and the other based only on respondents'set. A simulation study was performed to investigate the performance of the two estimators under various response situations. The simulation results showed that, under the True Response Model (TRM), the performance of the estimator allowing for nonresponse is superior when bias, mean squares and confidence level are used as performance measures. When the model is false (FRM), this estimator is inferior though slightly.