Four numerical experiments with different atmospheric forcing are considered to investigate a variety of transient responses of an upper-ocean density front similar to the Maltese front (south of Sicily) in an embedded mixed layer-ocean circulation model. Of particular interest is a case that stimulates forcing observed by O. M. Johannessen (unpublished report 1975): in May 1971, measurements of the Maltese front were being taken in calm weather, and then the wind suddenly increased to 10--15 m s-1 down the front. For this forcing the model result is consistent with the observations predicting steepening of the frontal interface, spreading of the surface isopycnals, and very little horizontal displacement of the surface front. However, the model predicts too much mixed-layer deepening and fails to predict the observed strengthening of the alone-front flow. Three additional experiments are considered to assess the effects of a change in the wind direction and the inclusion of a surface buoyancy flux. The inclusion of a diurnal surface buoyancy flux produces a very different response from a simulations with no buoyancy flux: when the surface buoyancy flux is negative, the mixed layer reforms at a shallow depth, resulting in much higher advecting velocities and subsequently larger displacements of the surface front. The negative buoyancy flux can effectively protect the subsurface structure with a shallow, stable surface layer; it also decouples the deeper frontal processes from the surface mixed-layer processes.