This research assesses the effect of employment to poverty reduction in Rwanda. For this purpose we investigate the micro-economic relationship between employment and poverty indicators. For the estimation of the ordered logit model, the Stata software was used to analyze data from EICV3 and to establish the relationship and the level of significance between poverty status as dependant variables and the household characteristics, employment variables as explanatory variables. They include the gender of the household head, marital status, education level, the status and sector of employment. From empirical results, households employed in wage- non-farm have high chance of being poor than those in wage non -farm. Factors that help household workers to avoid poverty are education. (from primary, secondary, higher education and vocational). The research suggests that modern technologies should be encouraged for households working in wage farm so as to increase productivity. Also, much emphasis should be put in investing for education that should increase workers capabilities, competence and thus increase productivity.