Purpose: This study determined the size-specific dose estimate (SSDE) of computed tomography (CT) examinations and derived mathematical expressions for dose output estimation and optimization in a teaching hospital in Ghana. Methods: Demographic and scanner output indices including CTDIvol and DLP for adult head, chest and abdominopelvic (ABP) CT examinations carried out at the hospital from 2018 to 2020 were retrieved from its CT scanner machine. Other indices such as antero-posterior (AP) diameter (DAP), lateral diameter (DL) and diagonal diameter (Ddia) of the patients’ bodies were measured on the mid-slice axial image using a digital caliper. The effective diameter (Deff) was then calculated as the square root of the product of the DAP and DL. The SSDEs were calculated as the product of the CTDIvol and the size-specific conversion factors obtained from the American Association of Physicists in Medicine’s (AAPM) Report 204. Regression analyses were performed to find the relationship between SSDE and the various parameters to derive mathematical equations for the dose estimations. Results: There were more female samples (n = 468, 56.3%) than males (n = 364, 43.7%) for each CT procedure. The SSDEs and size-specific diagnostic reference levels (SSDRLs) were: head (83.9 mGy; 86.9 mGy), chest (8.1 mGy; 8.7 mGy) and ABP (8.4 mGy; 9.2 mGy). The variations between CTDIvol and SSDEs for head (2.50%), chest (25.9%), and and ABP (26.2%) showed underestimation of radiation dose to patients especially in chest and ABP examinations if CTDIvol is used to report patient doses. The SSDEs of the chest and ABP CT examinations showed linear correlations with the CTDIvol. The estimated values could be used to optimize radiation doses in the CT facility. Conclusion: The SSDE and SSDRLs for head, chest and ABP CT examinations have been developed at a teaching hospital in Ghana. The SSDEs of chest and ABP examinations showed linear correlations with the CTDIvol and hence, can be calculated using the mathematically derived equations in the study.