This research looked at young children's ability to reason about the causal relations between events, and to make inferences based on temporal order information, a particular type ofmature and flexible temporal cognition. Children were presented with one oftwo variations ofa novel zoo paradigm. One version required reasoning about past events (search task) and the other required reasoning about future events (planning task). In studies 1 and 2, both versions ofthe zoo comprised five locations. In Study 1, both 3- and 4-year-olds failed the past and the future tasks, whilst 5-year-olds passed both. Four-year-olds performed significantly better on the future task. However, when the mode of response was changed in Study 2 to discourage serial searching, this task difference disappeared. Again, only 5- year-olds passed the two tasks, whilst 4-year-olds were not successful on either. In Study 3 when the number of events in the sequence was reduced from five to three, 4-year-olds were successful on this simplified version of the past task but not the future task. Children's difficulties on the future task persisted even when relevant cues were provided (Study 4). These results suggested that while 4-yearolds can make temporal-causal inferences about a simplified event sequence in the past, children below the age of five have difficulties doing so for future events. In Study 5, 4-year-olds still failed the future task even when given imaginative support in the form of an example of the goal state. Study 6 found that the difference between 4-year-olds' performance on the three- and five-location past task was not due to the presence of more than one possible correct answer. The results of these six studies are discussed in terms of the development ofchildren's temporal-causal reasoning abilities, and the wider issue of the nature of young children's temporal cognition.