This study was undertaken to assess the effect of ultrasound on human brain temperature in vivo. The investigation consisted of direct recording of intracranial temperature during color transcranial Doppler (TCD) sonography in a neurosurgical patient. The temperature was recorded from 3 thermocouples. One was implanted together with an intracranial pressure sensor into a surgically reduced intraparenchymal hematoma, the second was placed within the subdural space close to the temporal acoustic window, and the third was located extracranially at the outer surface of the temporal bone. Tympanic temperatures were also measured to give an approximation of global brain temperature. A 2.5-MHz transducer was used, and the system settings were as follows: spatial peak temporal average intensity = 234 mW/cm2 in B-mode at a maximum power of 32.3 mW and 2132 mW/cm2 in Doppler mode at a maximum power of 149.3 mW. Neither increase in the intraparenchymal brain temperature nor increase in the temperature at the bone/soft tissue interface was observed during 30 minutes of insonation. The ipsilateral tympanic temperature increased by only 0.06 degree C, and this value may be regarded as a measure of the overall increase in brain temperature. Passive cooling effect produced by the transducer, which was at ambient temperature, was found to reach the brain surface and to surpass any possible heating caused by the ultrasound. The results indicate that no noticeable increases in human brain temperature occur in response to ultrasound emitted by a color TCD device at high transmitter power settings within the diagnostic range.