本文以婦女勞動參與率為生育機會成本的指標，檢討晚近生育率下跌的需求面因素。我們使用1965年到1991年的資料首先就不分組迴歸模型估計婦女勞動參與率對生育率的影響，整組模型雖然有統計上的顯著性，但個別係數的影響效泉並不顯著，主要是受生育率單峰型態年齡分佈的影響。為反應年齡別生育力的自然現象與驗證婦女勞動參與率對生育率的影響是否因年齡不同而有所差異，我們就分組迴歸模型以OLS方法與GLS方法進行分析，指出婦女勞動參與率的提高對於晚近生育率的下跌的確有顯著的影響，尤其在生育力高峰點的年齡組上影響力更大。進一步利用Chow test檢驗分組與不分組是否有顯著差異，結果顯示婦女勞動參與率對生育率的影響在各年齡組之間的差異具有統計上的顯著性，分組迴歸模型是一個較好的分析模型。 如果婦女勞動參與率的上升是生育率下跌的重要原因，則有關人口老化的政策就值得我們進一步的檢討。生育率的下降是人口老化的主要原因，除了提升生育水準以避免過度的人口老化之外，鼓勵婦女就業也是我們經常強調的重點，在面對勞動力短缺與社會安全制度財務問題時尤然，則如何降低養育子女與勞動參與之間的衝突性乃是我們必須努力的方向。

Based on the concept of opportunity costs, this paper explores the effeds of women's labor force participation on fertility in Taiwan, from 1965 to 1991. As a first step, a pooled regression is applied to estimate the effects of women's age-specific labor force participation rates on age-specific rates of fertility. It is found that the model fitting itself is statistically significant but coefficients pertaining to the women's labor force participation are not. Subsequently a subgroup regression, using both OLS and GLS approaches, is applied to the estimation of coefficients. The GLS results show that while most of the coefficients are statistically significant, both the subgroup constants and labor force participation coefficients display an age pattern paralell to the age-specific fertility rates, suggesting that the opportunity costs might be important factor affecting the fertility at prime ages. A Chow test is constructed to compare the estimations of pooled and subgroup regressions. It shows that the subgroup regression fares much better than the pooled regression.