Biot’s model of a poroelastic medium has been shown to do well at predicting the result of passing an ultrasonic pulse through a cancellous bone specimen. The inverse problem of determining the parameters in the Biot model by matching the observed and predicted transmitted and reflected waves is not straight forward, however. Such problems, formulated as minimization problems, seem invariably beset by numerous local minima. In this talk the possibility of an indirect approach involving first solving a simple inverse problem, and then using this solution to attack the harder problem, will be considered.