Given that nancial protection in health is one of the objectives of the recent health system reforms in many developing countriesincluding Mexico, understanding clearly the mechanisms that lie behind the nancial vulnerability associated with health shocks is key for the success of past and forthcoming reforms. Using Mexican longitudinal data, I study the eect of breadwinners' health shocks on household consumption, measuring health shocks through changes in the capacity to perform Activities of Daily Living. I nd that health shocks to households' breadwinners are associated with signicant long-lasting decreases in non-medical per capita consumption, but health shocks to other household members have no such eects. When consumption depends on labor income, the economic cost associated with a health shock may lie less with direct out-of-pocket medical expenditures than with the diminished capacity to work. Therefore, providing health insurance to the previously uninsured sector of the population could potentially increase social welfare because of its consumption smoothing properties, although it would not necessarily provide full consumption insurance against health shocks. To illustrate this, I estimate the marginal per peso welfare gains paid in the premium of a full insurance against health shocks in Mexico and complement these calculations by using a standardized expected utility model, to compute the risk premia Mexican households would be willing to pay to reduce, either partially or entirely, households' risk exposure to health shocks.