Abstract: Based on the survey of tourists in Beijing，the tourist trip chaining behaviors were analyzed by Multinomial Logit model． Tourists travel attributes and personal attributes were considered to have a significant effect on the trip chain，and the allocation of time is also taken into consideration． The result demonstrates that six variables，including age，occupation， travel time，local tourists or not，be familiar with attraction or not，allocation of time，have significant impact on trip chain choice． The results show that the proposed model is of high accuracy，which can provide a theoretical basis for prediction of tourist travel．

Key words: tourism activities, trip chain choice, time allocation, Multinomial Logit model

摘要： 通过对在京旅游的游客进行调查，以游客的个人属性和出游属性作为影响因素，将游客对不同类型活动的时 间分配作为前提，对游客的出行链选择行为建立了Multinomial Logit 模型。模型分析表明: 年龄、职业、出游时间、是否本地游客、对景点是否熟悉以及游客对活动时间的分配等均对出行链选择有显著影响; 模型具有较高的精度，为游客出行预测提供了理论依据

关键词: 旅游活动, 出行链选择行为, 活动时间分配, Multinomial Logit 模型