This study investigates the effect of health shock on the household welfare and their comparative effects across regions and rural-urban settings in Ethiopia using Ethiopian Socioeconomic Survey 2013/14 and 2015/16 data. Two measures of health shocks were included in the empirical analysis: death of a household member and Illness of the household member. To analyze the data the study employed fandom effect model. The findings confirm that the effects of health shocks on consumption differ between consumption items, sub samples and health shock measurements. The illness of the household member had a positive but insignificant effect on all consumption categories and also the result shows illness of the household member has no significant effects on any of the sub-sample groups (regions, rural-urban settings and gender of the household head). The regression results cannot reject the hypothesis of consumption smoothing in the face of illness of the household member. On the other hand, results suggest that the death of the household member had a positive and significant effect on all consumption items and it has also a positive and massive significant effect on all consumption categories in Oromia region and on total and food consumptions in Amhara region. These findings suggest that there may be a need to introduce or strengthen social protection programs beyond the health sector to protect households from risk arising from health shock occurred in the household member.