The Medical Economics Survey-Methods Study was conducted under contract for the National Center for Health Statistics during 1975-76. The objective was to determine the cost-effectiveness of a variety of strategies under consideration for application in a national survey in 1977. Detailed data on health care utilization and expenditures were collected periodically from a panel of 691 Maryland households over a seven-month period. Issues tested through a random experimental design included the relative effectiveness of monthly versus bimonthly interviews and in-person versus telephone contact. Completeness and accuracy of information were determined through an extensive record check involving all providers and third party payers identified in the household check involving all providers and third party payers identified in the household survey. The results indicate that there are substantial deficiencies in the household reporting and that the household data for all types of medical services would benefit from the availability of record data in order to improve theri accuracy. The frequency of contact (monthly versus bimonthly) had little effect on the accuracy of reporting. Given the higher rate of attrition observed for the monthly procedures and their considerably greater cost, the bimonthly interval appears to offer advantages. In-person contact resulted in significantly better reporting compared with telephone contact in the Baltimore area, but no difference was found in more rural Washington County. Thus, the 10 per cent cost increase for in-person contact may well be justified by the improved accuracy of the data obtained for certain types of populations.