One-hundred-sixteen patients suffering from vascular headache (migraine or combined migraine and tension) were, after 4 weeks of pretreatment baseline headache monitoring, randomly assigned to one of four conditions: (a) thermal biofeedback with adjunctive relaxation training (TBF); (b) TBF plus cognitive therapy; (c) pseudomediation as an ostensible attention-placebo control; or (d) headache monitoring. The first three groups received 16 individual sessions over 8 weeks, while the fourth group continued to monitor headaches. All groups then monitored headaches for a 4-week posttreatment baseline. Analyses revealed that all treated groups improved significantly more than the headache monitoring group with no significant differences among the three treated groups. On a measure of clinically significant improvement, the two TBF groups had slightly higher (51%) degree of improvement than the meditation group (37.5%). It is argued that the attention-placebo control became an active relaxation condition.