Vessel blowdown and level-swell experiments are part of the thermal-hydraulics "Separate-effect" Tests (SETs) associated with BWR accident scenarios such as steam line break and loss of coolant. The predictive capability of TRACE (Version 5.0RC3), a USNRC-sponsored best-estimate thermal-hydraulics system code, has been assessed against the General Electric so-called "Small-vessel" tests (d = 0.30 m) and "Large-vessel" test (d = 1.19 m), as well as the Toshiba test (d = 0.49 m). Essentially, these experiments consisted in performing fast depressurizations of a tank partially filled with a quiescent steam-water mixture initially under thermodynamic equilibrium at 6.9 MPa and 0.7 MPa for the GE and Toshiba tests, respectively. As the system pressure drops, steam flashing occurs generating void in the bulk of the liquid and thus causing an upward displacement (or swell) of the liquid-vapor free interface, reversing to a downward movement (or shrink) when the flashing rate does not compensate the steam loss at the interface. The depressurization rates ranged between (dP/dt) = -0.47 MPa/s and (dP/dt) = -0.005 MPa/s. Comparisons between the predicted and experimental void fraction behaviors at various axial locations along the vessel were made, while considering various code modeling options. It has been found that satisfactory prediction can be achieved provided that the heat transfer role of the vessel wall is also taken into account. However, significant discrepancies were observed for the particular case with the highest rate of depressurization developing down to atmospheric pressure.