租買選擇和家庭組成都是影響住宅需求的最重要因素之一。若一個家庭決定選擇購屋，則是否需要貸款，貸款多少，又是另外一個必須同時決定的考量。雖然國內有一些相關文獻分別探討租買選擇、貸款需求、與世代組成對住宅需求的影響，但受限於國內現有資料，這些文獻一直都缺乏對上述這三個因素交互影響的聯合分析。本研究主要目的就採用1996年的家計收支調查資料，利用幾個不同層次的logit模型來分析這三個因素對住宅需求的交互影響關係。本文先採用條件式logit模型與多項式logit模型來進行估計，然後再利用Hausman 檢定來測試這二種模型的適當性。接著我們再建立一個巢式 logit模型，並利用充分訊息最大概似法(FIML)進行估計，我們發現FIML的估計結果符合巢式logit模型的要求。最後，在FIML的估計結果中，我們發現家戶總所得、家戶組成、與所得組成對於租買選擇及貸款選擇都有顯著的影響。 Tenure choice and household composition are important factors affecting housing demand. When a family decides to own a dwelling unit, it has to consider whether to set a mortgage from a bank. Thought there are some literatures in Taiwan considering the influence of tenure choice, household composition, and mortgage on housing demand, but there is little systematic analysis on the effect of the three factors on housing demand simultaneously. The purpose of this paper is using the Survey on Household Expenditure, Taiwan, 1996 and applying several levels of logit model to estimate the impact of the above three factors on housing demand simultaneously. A conditional logit model and a multinomial logit model are estimated first, and then the Houseman test is applied to check if these two models are appropriated for describing our data. Then building a nested logit model and applying the full information maximum likelihood (FIML) method to estimate the impact of these three factors on housing demand, we find that the model's requirement of the nested logit model is fulfilled by the FIML？？s estimated results. Moreover, we find that household income, household composition of generation, and income composition all have significant impact on tenure choice and housing demand.