Thermal ultrasound can be effective in increasing extensibility of collagen, thus aiding joint mobilization and stretching. In 1995, we reported on the rate of temperature decay following 3-MHz ultrasound in subcutaneous tissues. We repeated that study at 1-MHz frequency to see if the stretching window is different for deep muscle. Twenty subjects had two 23-gauge thermistors inserted 2.5 cm and 5 cm deep into their triceps surae muscle. We administered 1-MHz continuous ultrasound at 1.5 W/cm2 until the tissue temperature increased 4°C (vigorous heating). Immediately following the treatment, we recorded the rate at which the temperature dropped at 30second intervals. We ran a stepwise nonlinear regression analysis to predict temperature decay as a function of time following ultrasound treatment. There was a significant nonlinear relationship between time and temperature decay. At 2.5 cm, the average time for the temperature to drop each degree was: 1°C = 2:34; 2°C = 6:35; 3°C = 12:10: and 4°C = 21:14. At 5 cm, the average time for the temperature to drop each degree was: 1°C = 2:31, 2°C = 6:50: 3°C = 14:32; and 4°C = 27:49. Based upon prior research, thermal decay of 1-MHz ultrasound was slower than 3 MHz, and the deeper tissue cooled at a slower rate than superficial tissue following 1-MHz ultrasound. The data illustrated that the stretching window was open longer for deep-seated structures than for superficial ones.