This combined retrospective and prospective study describes trauma patients in an urban African Hospital and assesses whether use of trauma registries leads to improved clinical assesment. The Kampala Trauma Score (KTS) is assessed as an injury severity filter. The level of clinical assessment was defined by Model Rural Trauma Project (MRTP) trauma triage criteria. Trauma registries were filled out systematically for every alternate patient on arrival in the casualty department, and the patient status was recorded two weeks after admission. This retrospective study showed that 52% of the trauma patients were inadequately assessed. Amongst the deaths, 72.7% had been inadequately assessed (p value = 0.0193). Prospective data showed that injuries were most common amongst young males (72.7%), in and around the city following road traffic injuries (50.7%). The mean time between injury and arrival at the hospital was 0.363 hours (SD 0.331) and the mean hospital response time was 0.36 hours SD 0.245) The rate of inadequate assessment decreased significantly after the introduction of the registry (p value = 0.000). The case fatalities before and after the introduction of the registry was, however, not statistically significant. The KTS was found to be a reliable severity filter for injured patients, with a KTS score of less than 14 increasing the likelihood of death by at least three times. The results showed a cut off value of 12 by the ROC curve (0.8755; 95% CI = 0.8455 - 0.9055).