China's peasant workers, as a special group nurtured by urban-rural registration system, reached 274 millions in 2014. Studies on this massive group involve a great variety of aspects except the relationship between peasant workers 'income and their subjective well-being (SWB). This paper extends previous researches by systematically investigating the impact of income on peasant workers 'SWB based on the China Household Income Project survey (CHIPs) data. Besides, compared with the previous studies, this study contributes to the literature by taking into account the heterogenicity among peasant workers, discussing the different effects of income on SWB of peasant workers with different income levels, and further providing answers to why the impact of income on new generations of peasant workers 'SWB differs from that on their old counterparts.  The results of ordinal logit model reveal that: (1) The Easterlin paradox has emerged in peasant worker group, and it is not the family income per se but the remainder of family income after deducting necessary living expenses that positively affects peasant workers ' happiness, which indicates that raising peasant workers 'SWB is not confined to the satisfaction of basic physiological needs for the moment. (2) The urban-rural income gaps reduce all peasant workers ' SWB, but the reduction decreases with the rise of peasant workers ' income. However, income gaps among rural residents negatively influence low-income group 's happiness. (3) The effects of income on peasant workers 'happiness differ in two generation groups, specifically, income affects old generation group 's happiness more severely than that of new generation group. As far as income gaps are concerned, both the urban-rural and rural-rural income gaps have negative effects on old generation group 's happiness, but in terms of their new counterparts, the effect of the latter is not statistically significant. This conclusion illustrates that the new generation group does not treat rural residents as its reference group of relative income, so the views of Li Qiang and Knight on the choice of peasant workers ' reference groups do not hold true. (4) Beyond economical factors, non-economical factors such as health condition, level of education, marital status and the style of migrate also influence peasant workers ' SWB as well. This article proposes that it is of great importance to increase the earnings of the low-income group and narrow the urban-rural income gap so as to regulate the income distribution properly. In addition, refining the urban medical insurance system for peasant workers, exploring the off-site treatment mode of New Cooperative Medical System, enhancing peasant workers 'educational levels by vocational or technical training, as well as encouraging old generation groups to transfer nearby space should be the foci of governments ' attention. However, it is worth noting that the determinants of peasant workers ' SWB partially differ in two generation groups. The policymakers need not only to consider the similarities of two generation groups, but also to balance their differences to ensure the efficiency of policies.