A current focus in the prospective memory literature is the extent to which a prospective memory task (remembering to perform a future action) interferes with ongoing activities (defined in this study as lexical decision latencies). In the present study, participants had to detect one, two, three, four, five, or six prospective memory cues. Results showed no significant increase in lexical decision latencies with one or two targets; however, significant costs emerged with three or more targets. Furthermore, task interference showed a linear increase in task costs for word trial performance but not for nonword trial performance. Practical and theoretical implications of these findings are discussed.