This Digest describes four types of errors in mail surveys and summarizes the ways they can be reduced. Any one of these sources of error can make survey results unacceptable. Sampling error is examined through inferential statistics applied to sample survey results. In general, increasing sample size will decrease sampling error when simple random sampling is used, and other methods, such as cluster sampling and stratification sampling, can be used. Non-coverage error must be addressed; population lists generally do not cover all the population. A number of approaches are available to combat nonresponse error, and many techniques increase response rates. A fourth source of error is measurement error. The other types of error result from nonobservations or nonparticipation, but measurement error results from mistakes made by respondents. The increasing interest in mixed-mode surveys is likely to lead to more focused attention on measurement error issues. Reducing measurement error will be an important advance for this method of social science research. (Contains 18 references.) (SLD)