We report 2 experiments that assess factors potentially responsible for a proactive interference with the sensitivity of a mother's response to infant signals. Using a version of the "learned helplessness" paradigm, mothers' performance on a solvable task was assessed following pretreatments that involved exposure to an infant cry but that differed in the mothers' ability to exert control over the termination of the cry. Each experiment explored the role of varying attributions made by a mother in the development of, or the reversal of, the helplessness phenomenon. The first experiment addressed the question of whether a specific intervention (i.e., providing the mother an attribution for failure) can reduce the debilitating effect of prior experience with failure. The results indicated that the debilitating effects associated with previous failure were reversed for the 16 mothers assigned to the intervention group. 40 mothers participated in the second experiment, which varied attributions assigned to an identical cry stimulus (i.e., the cry was produced by an "easy" vs. a "difficult" infant). This experiment assessed the effect of varying attributions on the mothers' ability to terminate the cry. We found that mothers pretreated with inescapable cries and those receiving the experimental manipulation of attributing the cry to a "difficult" infant showed debilitated performance in stopping the cry when given the opportunity. We propose that models based on learned helplessness theory have value in the study of caregiver-infant relationships, in particular, caregivers' perceived and objective effectiveness in responding to an infant's signals.