SUMMARYThe role of latent-heat release in an explosively developing cyclone from the Fronts and Atlantic Storm-Track Experiment is examined within the potential vorticity (PV) framework. Using conventional synoptic analysis, piecewise PV inversion and numerical simulations, the development is described in terms of an upper-level PV (UPV) anomaly, a surface potential-temperature anomaly and diabatically generated PV anomalies in the lower and upper troposphere. The relative contributions of each anomaly to the cyclogenesis are investigated, along with the interactions between anomalies.In contrast to previous case-studies, the results of PV inversion and sensitivity tests reveal that the contribution from the surface thermal anomaly is much weaker than those associated with the upper-level and diabatically generated PV anomalies. A low-level PV (LPV) anomaly is generated rapidly in response to the UPV anomaly, and dominates the intensi cation of the system, but its associated circulation has only a modest feedback effect on the upper levels. However, diabatic reduction of PV in the upper troposphere is found to have important effects on the UPV anomaly. In particular, latent heating tends to destroy the UPV anomaly by inhibiting its downward penetration and reducing its horizontal extent.