Little is known about the impact of different types of headache on lost work time and work effectiveness in the United States. Estimation of the distribution and magnitude of the impact of headache in the workplace is necessary before workplace interventions can be developed. During 1993 and 1994 in Baltimore County, Maryland, study subjects were contacted by random-digit dialing and interviewed by telephone about their headaches. Headache diagnoses were assigned using International Headache Society criteria for migraine and episodic tension-type headache. Measures of workplace impact were derived based on self-reports of missing work because of headache and frequency and magnitude of reduced work level because of headache. Of the 13,343 respondents, 9.4% reported missing work more than rarely because of headache, 31% reported that their work level was reduced more than rarely by headache, and 9.2% reported that their work level was reduced more than 50% by headaches during work. In accounting for both actual lost workdays and reduced effectiveness at work, individuals lost the equivalent of 4.2 days per year because of headache. Of the 9922 annual estimated actual lost workdays because of headache, 57% were due to migraine and 43% were due to tension-type and other headache types. Of the 23,287 annual estimated reduced effectiveness workday equivalents, 64% were due to tension-type and other headache types, and 36% were due to migraine. Headache type, headache severity, and education level were each independent predictors of workplace impact of headache. Subjects with migraine headache were much more likely to report actual lost workdays because of headache, whereas tension-type and other headache types accounted for a large proportion of decreased work effectiveness because of headache. The results have implications regarding the control of indirect costs in the workplace because of headache, and on workplace-based treatment and prevention programs.