The present study tested the hypothesis that involving the partners of headache sufferers in a group cognitive-behavioural stress-coping intervention for chronic headaches would enhance treatment outcome. A treatment program in which partners were involved (PI) was compared with a standard treatment program in which partners were not included (NPI) and a no-treatment control condition (NTC). The NPI condition was effective compared to the NTC condition, but there was less evidence that the PI condition was effective. No significant differences were found between the PI and NPI conditions at posttreatment or follow-up. These results are considered in terms of the usefulness of involving partners in group headache-treatment and the manner of their involvement. In addition, the possibility that partner involvement may actually detract from treatment effectiveness is discussed in terms of its impact on motivation and personal responsibility for change.