The limitations of some commonly used model-based methods of diagnosing extratropical cross-tropopause transport in tropopause folding events are revealed by performing forecasts of a case study using the Met Office Unified Model including an idealized passive tracer. Evidence is presented that the transport which occurs in parameterized convection and turbulent mixing can lead to significant cross-tropopause transport. Transport from the stratosphere to the troposphere by parameterized processes (across a potential vorticity defined tropopause zone) was found to be dominated by convective transport. Transport by parameterized processes accounted for a maximum of 38% of the tracer transported by both resolved and parameterized processes in a control forecast (integrated across the model domain); alternatively, transport by parameterized processes led to a maximum increase of 62% in the transported tracer above that transported by resolved processes only. The importance of these processes depends on the synoptic system under consideration and local effects can greatly exceed the domain average. These processes are neglected by methods of diagnosing cross-tropopause transport based on trajectories calculated using the resolved scale winds or based on the nonconservation of potential vorticity. Estimates of transport are highly sensitive to model horizontal and vertical resolution and to the artificial diffusion required for model stability. The transported tracer was 78% greater at the end of the forecast in which most transport occurred compared with the forecast in which least transport occurred. Finally, transport by parameterized convection was found to be even more important when considering transport from the stratosphere to below 700 mb, a transport measure used in previous studies.