伴隨著資訊社會的高度化，於市中心導入「停車導引資訊系統」，將是未來都市交通建設的整體策略中，一項必要且重要的發展方向。然而目前的研究大多仍僅止於探討其硬體面或技術面，至於「改善的效果」與「如何影響停車行為」的量化測度則較少觸及。由於國內尚未有實施停車導引資訊系統的地區，故本研究乃採個體選擇模式中的「敘述性偏好法」來掌握賀駛者之行為模式與偏好。為驗證本方法之可用性、有效性及合理性，以處理臺南市中心區購物停車需求行為之研究為例為行實證分析。實證結果確認了「資訊提供」可提高停車選擇「效率性」與「機會性」的效果，且駕駛者對於資訊提供具有「迴避型風險」的傾向。

　　　　　Following the fast development of information society, it is necessary and important to develop the direction for adopting "Parking Guidance Information System (PGIS)" in CBD for the strategies of urban constructions. Nowadays, most of studies merely investigated the hardware and techniques of PGIS and rarely discussed it's "improvement effect" and "how it influence parking behavior". Since PGIS wasn't implemented in Taiwan yet, "Stated Preferences" of discrete choice model would be used to capture the drivers; behavior and preferences. To identify the feasibility, effectiveness, and reasonableness of the model, we executed a case study in Tainna CBD to measure driver's parking behavior. The effects that information service promoted the efficiency and opportunity of parking choice was identified. Meanwhile, we found that drivers tended to avoid the risk of information errors.