Two experiments investigated the role of inflections in verb learning. In Study I, 3- to 5-year-olds with typical language development were asked to extend novel verbs to new instances. They heard the verbs inflected with either -ed or -ing and were given a forced choice between events that maintained either the activity or the result of the original event. The younger children selected events according to the verb inflection: same-activity events for -ing and same-result events for -ed. Older preschoolers chose same-result events throughout. Study II was conducted to investigate the nature of this causal bias. A group of 4- to 5-year-olds with specific language impairment completed the same verb extension task. They were equivalent to the older Study I children in age and IQ but were at lower language levels than the younger group. Children in the SLI group used neither the inflectional strategy nor the same-result strategy. Findings from the two studies point to a developmental period during which children treat inflectional cues as reliable guides to verb meaning. The discussion focuses on the rise and fall of such inflectional bootstrapping and the linguistic character of the same-result bias that replaces it.