The employment rates of prime-aged males in Taiwan have been decreased after 1994. Since the labor supply of prime-aged males play the major roles in family incomes, these large and persistent increases in non-employment rates would raise the number of families which bore the substantial shocks in family income. The purpose of this paper is, then, to investigate the labor supply responses of family members to mitigate their family’s income loses when their male household head experienced non-employment. To examine the labor supply responses of family members in different time span, we use a pseudo-panel data set constructed from 1993-2005 Manpower Utilization Survey. There are two measurements of labor supply: one is employment or not employment, and the other one is the hour of work given employ. The fixed effect Logit model and fixed effect OLS model are used to run a empirical regression analysis. More, the model of Heckman’s two stage estimated method is used to estimate the fixed effect OLS model. We separate the family condition into single-parent family and two-parent’s family, and discuss if the male household head was displaced unemployed or not. Also, we want to see the employment and the hours of work changes between wives, children or other family members in the current or next period when the male household head was unemployed, and to observe if there is added worker effect in Taiwan’s family. The empirical result shows that the added worker effect existed apparently if we discussed it based on employment. For two-parent’s family, whether husband was displaced unemployed person or not, wife and other family members increased their labor participation in the next period after husband unemployed. For children, when father was general unemployment, the added worker effect existed in that period. And if father was the next displaced, the added worker effect appear in the next period. However, no mater what the male head of family was a general or displaced unemployed person for single family; the children had no response of labor participation in current or later period. But in other family members, the added worker effect existed, and they would have noticeable response to the male household head of family unemployed in later period. Finally, in term of by hours of work, we found that all family members hadn’t any response of the added worker effec.