The Sudakov veto algorithm for generating emission and no-emission probabilities in parton showers is revisited and some reweighting techniques are suggested to improve statistics by oversampling in specific cases.