This retrospective review examined the cause, level of pathology, onset of symptoms, time taken to treatment, and outcome of 19 patients with cauda equina syndrome (CES). The minimum time to follow up was 22 months. Logistical regression analysis was used to determine how these factors influenced the eventual outcome. Out of 19 patients, 14 had satisfactory recovery at 2 years post-decompression; 5 patients were left with some residual dysfunction. The mean time to decompression in the group with a satisfactory outcome was 14 h (range 6–24 h) whilst that of the group with the poor outcome was 30 h (range 20–72 h). There was a clear correlation between delayed decompression and a poor outcome (P = 0.023). Saddle hypoaesthesia was evident in all patients. In addition complete perineal anaesthesia was evident in 7/19 patients, 5 of whom developed a poor outcome. Bladder dysfunction was observed in 19/19 patients, with 12/19 regarded as having significant impairment. Of the five patients identified as having a poor overall outcome, all five presented with a significant sphincter disturbance and 4/5 were left with residual sphincter dysfunction. There was a clear correlation between the presence of complete perineal anaesthesia and significant sphincter dysfunction as both univariate and multivariate predictors of a poor overall outcome. The association between a slower onset of CES and a more favourable outcome did not reach statistical significance (P = 0.052). No correlation could be found between initial motor function loss, bilateral sciatica, level or cause of injury as predictors of a poor outcome (P > 0.05). CES can be diagnosed early by judicious physical examination, with particular attention to perineal sensation and a history of urinary dysfunction. The most important factors identified in this series as predictors of a favourable outcome in CES were early diagnosis and early decompression.