The ability to recall is something that most intact adults take for granted. For much of the last century, this feature of mental life was not considered to extend to very young children. There now is evidence that 1- to 2-year-olds are able to recall specific events after delays of several months. Over the short term, 1- to 2-year-olds' recall is affected by the same factors that affect older children's recall; it is not clear whether similar effects are apparent over the long term. Moreover, although age-related increases in long-term recall are assumed, there have been few empirical tests of the question. We examined recall by 14- to 32-month-olds for events experienced at 13 to 20 months. Using elicited imitation of novel multistep event sequences we examined effects of (a) delay length, (b) age at the time of experience, (c) temporal structure of events, (d) mode of experience of events, and (e) availability of verbal reminders, on long-term recall. Participants were 360 children enrolled at 13 (n = 90), 16 (n = 180), and 20 (n = 90) months. All of the 13-month-olds and half of the 16-month-olds were tested on 3-step event sequences; all of the 20-month-olds and half of the 16-month-olds were tested on 4-step event sequences. Within each age and step-length group, equal numbers of children were tested after intervals of 1, 3, 6, 9, and 12 months (n = 18 per cell). Children were tested on a variety of sequence types. For half of the events, imitation was permitted prior to the delay; for the other half, children were not permitted imitation. At delayed testing, children experienced a recall period during which they were cued by the event-related props alone, followed by a period in which recall was cued both by the event-related props and by verbal labels for the event sequences. Within step-length groups, the length of time for which older and younger children showed evidence of memory did not differ. Nevertheless, when the children were prompted by the event-related props alone, there were age-related differences in the robustness of children's memories (as indexed by higher levels of recall for older children relative to younger children). When the children were prompted by the props and by verbal labels for the event sequences, at the longer retention intervals, there were age-related differences in the robustness of children's memories and in the reliability with which recall was evidenced (as indexed by the larger numbers of older children evincing recall). Age-related effects were particularly apparent on children's ordered recall. Across the entire age range, the children were similarly affected by the variables of sequence type, opportunity for imitation, and verbal reminding.