African emergency medical services (EMS) systems are inadequate, thereby necessitating its selective use. This study aims to investigate differences in mode of arrival to the Emergency Centre (EC) at Komfo Anokye Teaching Hospital in Kumasi, Ghana by acuity, injury and referral status. Methods: A cross-sectional survey was conducted in the EC at the Komfo Anokye Teaching Hospital (KATH) in Kumasi, Ghana, in 2011. A survey was administered to all patients triaged to the EC. Patients were excluded if they were under 18 years of age, unable to communicate in English, Twi, or Fante, had altered mental status, or were deceased. Data were inputted into an excel spread sheet and uploaded to SPSS. Descriptive statistics were computed. Inferential statistics were performed testing for differences and associations between modes of arrival and acuity level, referral and injury status. Results: Of the 1004 patients enrolled, 411 (41%) had an injury-related complaint, and 458 (45.6%) were inter-facility transfers (“referrals”). 148 (14.8%) arrived by ambulance, and 778 (77.6%) non-ambulance (38% private cars, 38% taxis). 67 (6.7%) were triaged as Red, 276 (27.5%) as Orange, and 637 (63.4%) as Yellow (highest to lowest acuity). Ambulance arrival was positively associated with a higher triage score (OR: 1.53). Patients referred from other facilities were almost twice as likely (OR 1.92) to arrive at the KATH EC via ambulance than those not referred. Patients with injuries and higher acuities patients were more likely to be transported to KATH by ambulance (OR 1.86 and 1.87 respectively). All results are highly statistically significant. Conclusion: Although a minority of patients were transported by ambulance, they represented the most acute patients arriving at the KATH EC. Given the limited availability of EMS resources and ambulances in Ashanti, selective ambulance use appears warranted and should inform prehospital care planning.