Objective. To determine predictive factors for radical treatment (nucleolysis or surgery) after in-hospital conservative management of low back pain with sciatica (LBPS). Patients and methods. A standardized form was used to collect data on 134 patients admitted for conservative treatment of LBPS. Subsequent radical procedures were recorded 11 to 24 months after discharge. Results. Forty-seven patients required radical treatment after discharge. Significant risk factors for radical treatment in the univariate analysis were taller stature, use of a lumbar support, more preadmission epidural injections, a positive straight leg-raising test, and a disk herniation diameter of at least 50% of the spinal canal diameter. Protective factors were onset within the month preceding admission and normal range of motion of the lumbar spine. In the multivariate analysis, symptom duration longer than one month, use of a lumbar support prior to admission, and a positive straight leg-raising test were associated with radical treatment. A positive straight leg-raising test was the only significant clinical risk factor in the subset of patients investigated by computed tomography (CT). When CT findings were added to the model, only size of the herniation was significant. Conclusion. Sixty-five percent of patients admitted for conservative treatment of LBPS do not receive radical treatment during a mean follow-up of 18 months. Several factors are associated with the likelihood of radical treatment.