Thirty-two patients with chronic low back pain were treated three times at weekly intervals with rotational manipulation. Patients with femoral or sciatic root pain were included provided they did not exhibit root compression signs. Background therapy of codeine phosphate was administered throughout. There was a significant increase in spinal flexion measured clinically during the three-week period of manipulation followed by a significant decrease in the three-week period after manipulation. The first week of manipulative treatment was more Painful than the corresponding week in the control group but in the second and third weeks there was less pain in the manipulated group. Pain scores were reduced to a significant degree within four weeks of starting treatment only in the group manipulated in the first treatment period. Patients benefitting subjectively from manipulation were more likely to be older and to have had symptoms for a shorter period than those not deriving benefit. The age of onset of symptoms was significantly later in the responders.