Efficient experimental designs offer the potential to reduce required sample sizes, or to reduce confidence intervals for parameters of interest, in choice experiments. Choice experiment designs have typically addressed efficiency of utility function parameter estimates. The recently developed concept of C-efficiency recognises the salience of willingness to pay estimates rather than utility function parameters in studies that seek to put money values on attributes. C-efficiency design benefits have been illustrated in a theoretical context, but have not been tested in applied settings. This study reports a choice experiment field application that used initial responses to update statistical designs to maximise C-efficiency. Consistent with theoretical predictions, the revised design delivered significant reductions in the variance of willingness to pay estimates, illustrating that C-efficient designs can indeed decrease costs of choice experiments by reducing required sample sizes.