The issue of illness reporting in modelling demand for health care in low- and middle-income countries can be handled according to either of two conceptually-different constructs: (a) considering illness reporting behaviour as endogenous to demand; or (b) considering demand itself as the outcome of a sample selection phenomenon. In this paper, we take the second viewpoint and estimate the demand for medical care with an estimator that uses Heckman-type. Empirical estimates based on household survey data from rural Burkina Faso suggest that there are some implications of illness reporting behaviour for modelling the demand for medical care.