在數位時代來臨的今日，利用數位媒體或科技將虛擬的遊戲世界與現實的世界連結整合的隨境遊戲，乃是未來趨勢上相當重要的一種行銷手段。然而目前學界上卻沒有確切關於隨境遊戲的一套設計方法。有鑒於此，本研究先蒐集相關的理論與研究，進而發展出隨境遊戲模型與刺激閾模型。利用這兩個模型設計一個以推廣流浪動物TNR為議題的隨境遊戲—Trap paw。在Trap paw中，玩家可以透過各種不同的方式與情境在現實世界裡面尋找線索來進行遊戲。之後，研究者採用半結構式的深度訪談法，訪談了六位不同身份背景的玩家，以探詢他們對於遊戲的回饋與建議。綜合此訪談結果與創作經驗，除了修正Trap paw的遊戲內容外，也提出了關於使用上述兩個模型進行創作的幾點建議。 在實作之後，發現這兩個模型宛若隨境遊戲的基石般穩固、實用性也很高，沒有需要特別修改之處。只不過若要應用在行銷活動式隨境遊戲設計上的話，針對隨境遊戲模型的使用可提供三點建議：（1）考慮先從明顯易見的空間擴增發想遊戲架構。（2）嘗試創造可見的魔法圈界。（3）舉辦效果良好的慶典式遊戲活動。至於使用刺激閾模型的部分，則建議：將「權變因子」這個因素作為主動促使個體逆轉成容易接受遊戲的行為導向狀態；將「厭膩」與「挫折」這兩個因素作為尋找個體已經是處於行為導向狀態的情境。透過這兩種面向的思考，以求在正確的時間與場合讓個體接觸到遊戲，使其容易接受遊戲而成為真正的玩家。 / In this digital marketing generation, the pervasive games which connected the game world with the reality by ubiquitous computing or digital medias get more and more attentions as a method for marketing. However, there are still not specific design theories or methods of the pervasive games. For that I developed two models that can be used to design the pervasive game: pervasive game model and liminal model by the collections of concerned theories and studies. By the pervasive game model and the liminal model, I created a pervasive game which is about promoting TNR-—Trap paw. In the Trap paw, players can find clues in a wide range of ways and in the different scenario out of their imagines. Qualitative interview is adopted as a research method and I interviewed six people to look for the feedbacks and suggestions about the game. Combined with the outcomes of interviews and my personal experience, I modified parts of the game. In the end, I give some suggestions about creating a pervasive game by using these two models. After completing the game, I find out that the two models are not only stable as the foundation stones of the pervasive games, but they can be applied widely. No further modifications are needed. However, when applying the pervasive game model to the field of promoting campaign, there are three suggestions: (1) begin with the space expansion (2) try to make some visible or tangible magic circles (3) through some gaming ceremonies. As for using liminal model to developing a game, I suggest that “contingent events” can be used to be an active factor to reverse individuals from telic state into paratelic state and concerning “frustration” and “satiation” to be factors to define whether an individual is in the paratelic states. Through these kinds of thinking, game might contact to the target audience in the right time and space. In this way, it’s easier for them to embrace the game and become real players.