Ninety-nine meta-analysts returned their questionnaire. The largest number of respondents identified their discipline as psychology (n=40), followed by medicine (n=18), education (n=15), nursing (n=6), and epidemiology (n=4). Other fields included in the final sample were organizational behavior, communications, management, health services research, dentistry, biostatistics, social work, and public health. Participants reported having conducted or been involved in from 1 to over 30 meta-analyses, with a median of 3. The questionnaire asked respondents to rate how helpful each of 14 items that addressed "policies and products" would be in facilitating the cumulative synthesis of knowledge through meta-analysis. A 5-point response scale was used. Although a number of differences were found between medical and social/nonmedically related health scientists, the greatest degree of overall enthusiasm for both groups was found for the need to educate journal editors and primary researchers about information that should be reported in an empirical study. The respondents were in general agreement that a consensus should be developed and disseminated on minimum standards for published meta-analyses. A number of the other proposals were less popular. The authors note that since the conclusion of the survey, many of the proposals have been implemented. 3 references and 3 tables