Hydralazine is an antihypertensive drug which theoretically could increase tumour temperatures during hyperthermia via reduction in tumour blood flow from a vascular 'steal' phenomenon. Doses that are therapeutically effective in reducing blood pressure in hypertensive patients would probably cause postural hypotension and other side-effects in normotensive patients beyond the hyperthermia treatment session, however. This study was designed to evaluate whether hydralazine, when administered at a safe dose for normotensive patients (0.125 mg/kg, i.v.) would be effective in increasing tumour temperatures during hyperthermia. The working hypothesis was that hydralazine at a dose of 0.125 mg/kg would be effective in raising tumour temperatures during hyperthermia treatment with minimal change in blood pressure. Fourteen human and five canine subjects were given hydralazine (0.125 mg/kg, i.v.) at the midpoint of a hyperthermia session. Temperatures and blood pressures were monitored before and after drug administration. Although hydralazine resulted in slight reduction in blood pressure, it was ineffective in increasing tumour temperatures in human patients (average maximum rise in median temperature was 0.26 +/- 0.32 degrees C). In canine subjects the same dose of hydralazine was effective in reducing blood pressure in four of five subjects studied (mean maximum drop was 22.7 +/- 4.1 mmHg) and the median temperature rose 0.8 +/- 0.7 degrees C. In the canine subjects the greater the decrease in blood pressure, the greater the increase in temperature. These results suggest that a rise in tumour temperature induced by hydralazine is dependent on creating a drop in blood pressure. Future studies in this laboratory will include tumour blood flow manipulation with antihypertensives which have a shorter half-life and a titratable effect. Using this approach, hypotension, which seems to be required to raise tumour temperature, will be more controllable in terms of magnitude and duration.