In a prospective, non-randomized study 40 athletes with contusion or distension injuries to the thigh or the calf muscle were followed with tests of range of motion (ROM) of knee or ankle joint, test of serum creatine kinase (CK) and ultrasonography of the injury until completely recovered. An experimental group of 19 injuries where subjects received treatment with application of a maximum compression bandage within 5 min (mean = 2 min) of the injury was compared to a control group of 21 injuries where subjects were treated with rest and elevation only, and in some cases non-maximum compression after 10-30 min. No significant differences were noted with respect to time to complete subjective recovery, ultrasonic size of the injury or time to normal findings on ultrasound between treatment and control groups. Strain injuries, although showing a tendency to be smaller in size, took a longer time to complete recovery than contusion injuries (mean +/- SD = 26 +/- 22 days and 19 +/- 9 days, respectively, P = 0.02). Diagnostic CK values and reductions in ROM were not correlated to the severity of the trauma, while ROM showed weak correlation to the sonographically measured size of the hematoma (r = 0.42: P < 0.01). Injuries displaying a circumscript anechoic, low-echogenic or mixed lesion at the diagnostic ultrasound investigation normalized more slowly (P = 0.001) and took longer to complete recovery (P = 0.001) than injuries with diffuse hyperechogenic lesions. We conclude that in this study the application of a maximum compression bandage within 5 min of a muscle trauma did not significantly reduce the size of the hematoma nor significantly shorten the time to complete subjective recovery compared with no immediate treatment. The diagnostic ultrasound investigation was valuable in predicting the severity of the trauma.