The contribution of the present paper is threefold. First, we formally test whether the effect of calorie deprivation on wages is more significant/higher for the lower quantiles of workers. In the extant literature this is established through non-linear terms in the wage equation. A more satisfactory method of doing this is through quantile regressions. Second, the quantile regression approach helps us identify the exact group for which the poverty-nutrition trap holds. The extant literature is unable to establish whether there are systematic differences across different quintiles in the response of productivity/wages to nutrition. The present paper addresses this lacuna. Third, we are able to establish a critical wage level for which the PNT trap hypothesis holds. For wages higher than this the hypothesis does not hold. We then argue that this value of the wage rate should set a floor for any minimum wage for agricultural labourers.