This paper reports the findings from an experimental study involving thirty three secondary school students (mean age = 15.5 years) in spatial visualization (SV) training through an interactive desktop virtual environment spatial trainer (iDVEST). Stratified random sampling was used to assign students into two experimental groups and one control group. The first experimental group trained in interaction-enabled DVEST (i-DVEST), the second experimental group trained in animation-enhanced DVEST (a-DVEST), and the control group received conventional training. A multi-factorial pretest posttest design procedure was used and data were analyzed using 2-way ANCOVA. Participants trained in i-DVEST made the highest improvement gain in SV, followed by those trained in a-DVEST, and the control group achieved the lowest improvement gain. In general, male participants achieved higher SV improvement gain compared to female participants. Interaction effect between method of training and gender was observed indicating that male students tended to benefit more when trained in i-DVEST, moderately in a-DVEST, and poorly in conventional method. On the other hand, female students seemed to benefit from training irrespective of the method used. (Contains 3 figures and 8 tables.)