In many cases of whiplash injury symptoms persist and do not respond to treatment. There is uncontrolled evidence to suggest that intracutaneous injections of sterile water might help. Since that route may be unacceptable to patients the subcutaneous route is used in the randomised trial reported here. 40 patients with whiplash syndrome, mean age 46 years (24-73) were given subcutaneous injections of 0.3-0.5 ml sterile water or saline over tender and trigger points in the neck and shoulder. A maximum of three treatments were given during the first two months of the study and the patients were followed up for 8 months. The accidents had occurred 4-6 years previously. X-ray examinations revealed no traumatic spinal lesions. Neck mobility and pain levels were evaluated by a physiotherapist immediately before and after the first treatment and after 1, 3, and 8 months. After 3 months, the mean total mobility of the cervical spine had increased by 39 degrees in the sterile water group and 6 degrees in the saline group (p < 0.05). Minimum and maximum levels of pain in the weeks just before treatment were evaluated by a visual analogue scale from 0 to 10. After 3 months the minimum pain level had fallen from 2.2 to 1.4 in the sterile water group but was not reduced in the saline group (p < 0.02); the maximum had fallen from 8.1 to 3.8 in the sterile water group and from 8.3 to 7.5 in the saline group (p < 0.001). After 3 months, 19 of 20 patients in the sterile water group assessed their condition as generally improved but only 6 in the saline group felt that they had got better. After 8 months there were still significant differences for minimum pain score and for mobility but not for maximum pain or for self-assessment of improvement.