本文目的為應用多變項線性迴歸廣義估計方程式方法（GEE Method Multiple Linear Regression），以兩項銀行產出變數：存款，放款等評估大型中國大陸銀行分行在地理分布及空間型態上績效之差異與演變。為求數據的一致與連貫性，本文選用中國大陸四大銀行之一且2012及2013年連續兩年排名全球第10大銀行:中國農業銀行為例，研究期間為1990年至2008年，2008年後省市數據不再公布。本文檢定出該行渤海區域與珠江三角洲等兩區域之分行績效差異並不顯著，僅次於績效高居六大區域首位之長江三角洲。分行績效不僅反映所在區域經濟發展程度之差異，亦呈現位於沿海分行績效普遍優於內陸分行，同時分行與總行間地理距離遠近有影響分行績效之傾向。該行在經營效率上，特別是位於中部，西部，東北部等三大區域分行距離總行北京較遠，績效明顯落後，勢需再加強提升，因應未來中國大陸銀行業加速開放後內外之競爭壓力。 This paper aims to assess the performance of China’s banks by their geographical distribution of loans and deposits , with the application of Generalized Estimating Equation (GEE) method’s multiple linear regression. We chose mainland China one of the four largest banks also is 2012 and 2013 for two consecutive years ranked the world''s 10th largest bank, Agriculture Bank of China, with data from 1990 to 2008 as the sample bank. We identified that branches’ performances in the Yangtze River Delta ranked first and were followed by the Pearl River Delta. Different than that of the previous literature, the results further indicate that branches in economically well-developed regions were not necessarily better-performed during this period. The empirical evidence also implies that distance impacts performances between branches and their headquarters. We emphasize that China’s banks benefited greatly from the support of various government policies. The management efficiency of the ABC in the other four regions particularly the Bohai Sea and the Central regions with larger GDPs, needs to be strengthened with further deregulation of China’s banking industry and financial markets to domestic and international competition in the coming years. Overall the evidence provides insights into bank-specific management in efficient allocation of scarce resources and could contribute to inform government policy, as well as to improve managerial performance and research issues on this regard.