The relative effectiveness of four different training conditions on the attainment of conservation of quantity was examined in a 5-Group design. Subjects were 75 nonconserving preschoolers, ranging in age from 3.10 to 5.0 years. Treatment consisted of five conditions: (1) pretense play training, (2) direct conservation training, (3) a combination of pretense play and conservation training, (4) mere exposure to the conservation tasks without the benefits of instruction, and (5) a control condition which provided only conservation pre-tests and post-tests. The results of the two conservation posttests indicated that a combination of pretense play and conservation training was the most effective condition, followed by the conservation training and then by the pretense play training. The data also indicated that the effects of conservation training were less stable than those attained with pretense play training, and that the former tended to decrease with time while the latter showed a corresponding increase over time. (Author/SS)