A comparison is made between the General Health Questionnaire (GHQ) and the Symptom Checklist (SCL) as psychiatric screening tests in community-based research projects. Both are shown to correlate equally well with independent clinical assessment, and the differences between them mainly reside in the form of their response scales. The GHQ works best as a screening test, since it has fewer false positives associated with its use, but it may miss those with long-standing disorders. The SCL tends not to miss long-standing disorders and furnishes diagnostic sub-scales if these are required. Both tests function better with men than with women and with whites than with blacks, but neither is affected by social class or age of the respondent. The study revealed high correlations between the symptoms of anxiety and depression, and indicated some possible differences between the symptom clusters seen in whites and in blacks.