This study examined the hypothesis that the response rate in the U.S. Census 2000 could be increased by explicitly offering respondents an alternative to using the paper form. A random sample of 10,494 households received special Census 2000 mailings. Half of these households were offered the alternative of responding via the web, while the other half were offered a telephone alternative. The offer of these alternative response modes boosted response rates by more than two percentage points over households that were not offered a response mode alternative. Half the mailings included an insert that held a telephone calling card incentive, which was activated if the household chose to use the web or telephone alternative. The insert holding the calling card greatly increased responding via the alternate response modes, but this increase was more than offset by decreased responding via the paper form. The implications of the results for future decennial censuses are discussed.