The results obtained from several forging and extrusion tests at room temperature and low speed are compared with theoretical results, and the following conclusions emerge. 1. (1) The working pressures for the compression of soft aluminium cylinders between rough flat dies, for the compression of soft aluminium annuli in a container between lubricated or rough dies and for the steady-state lubricated extrusion and piercing of hard aluminium and copper cylindrical billets, are 10–20 per cent below the calculated upper bound. 2. (2) Calculated upper bounds agree qualitatively with the working pressures obtained during some non-steady extrusions and piercings of soft aluminium, hard aluminium, copper and soft brass. 3. (3) The steady pressures divided by the mean yield stress in combined bar and tube extrusions from cylindrical billets and in tube extrusions from hollow billets, together with those in ordinary extrusion and piercing, were found to be almost entirely a function of the total reduction in area, when the tools are lubricated. 4. (4) The working pressure in opposed extrusion-piercing-forging is considerably lower than that for a one-way extrusion-forging. The working pressures in opposed extrusion-forging with two dies and in opposed piercing-forging with two punches were lower than that in corresponding one-way working, using a die or punch of the larger orifice when the length of billet is moderate. In symmetrically opposed working, the pressure exceeds that in one-way working when the billet length becomes very small. When the material flow through one orifice is interrupted midway in an opposed forging, the working pressure thence follows that of the corresponding one-way working. 5. (5) The deformation of hard aluminium and copper annuli compressed between lubricated flat dies was as predicted from the most suitable velocity field. The speed of efflux of material through two separate orifices was affected by the frictional resistances acting on the material rather than by the difference in the areas of the orifices when the billet length was moderate. But, as the length was reduced, the speed from the wider orifice tended to increase. These results were in qualitative accordance with those derived from plane-strain analysis. 6. (6) The internal flow pattern observed on the meridian section of a lead billet as photographed through a glass plate was in qualitative agreement with the corresponding most suitable velocity field and was very similar to that of the comparable plane-strain working. 7. (7) The main types of defect found in the various extruded and extrusion-forged products were skin inclusion, cavity formation and internal cracking. They were predicted quantitatively and qualitatively from the velocity field of the least upper bound. The inclusion was prevented by using a rough die in contact with the material surface from which the skin was unfolded. Cavity formation and internal cracking was thought to be prevented by restraining the material flow through the orifices. The onset of the latter defects was retarded by intermediate annealing. 8. (8) The internal pressure acting on the container during extrusion-forging and coining was found to be lower than the mean working pressure within the range of conditions tested. The pressure acting on the projection of the coining die was in excess of the mean working pressure, the difference being larger for higher projections and for thinner slug thicknesses. The residual pressure between the material and the container after extrusion or closed-die coining was found to be nearly equal to the uniaxial yield stress of the material.