The current method for creating tactical logistics estimates in the Army inadequately incorporates historical data on the actual consumption of supplies. The automated-logistics-planning system described in this thesis addresses this deficiency. The program developed in this research produces general estimates for selected supply items by referencing equations and variables from current Army planning documents and performing the necessary calculations. The program uses reasoning to identify previous operations which are analogous to the current operation. Separate criteria are used to identify the strongest analogies to the current operation for each of five categories of supply items. Information contained in the historical records of the three stongest analogies in each category is used to revise the general estimates. The revised estimates are hopefully more accurate in predicting actual supply requirements for the current operation than the estimates generated by formula alone. Theses.