It has been well established that life satisfaction is related to perceived social justice. However, current theories provide contrary assumptions on the direction of the influence. In this research, we use data from two longitudinal surveys collected in China to test the reciprocal relations between life satisfaction and perceived social justice over time. With a random intercept cross-lagged panel model, we disaggregate the between-person effect and the within-person effect of the relationship. To specify the conditions of the effect, we consider income levels as the moderator. Study 1 (N = 119) showed that on the between-person level, life satisfaction, and perceived social justice are positively correlated. On the within-person level, cross-lagged effect results showed that an individuals' deviations from their expected score in life satisfaction predict deviations from their expected perceived social justice at the next time point, while deviations from expected perceived social justice does not predict subsequent deviations from expected life satisfaction. In study 2 (N = 637), we divided participants into three groups based on their household income and conducted a multiple group analysis to test its moderation effect. We found that the between-person correlation of life satisfaction and perceived social justice is not moderated by income level, and it is significant in all the three groups. However, the within-person cross-lagged effect is moderated by income level, and the effect of life satisfaction on perceived social justice only exists in the low income group. This research confirms the unidirectional relationship between life satisfaction and perceived social justice across time, and clarifies the effect in different levels and income groups, providing new insights on the formation of justice perception. It is recommended that future studies apply experimental designs to reach causal effects and explore more possible moderators and mediators.