This study was conducted to explore how the level of imagery in a radio advertisement affects the allocation of cognitive resources to encoding the message into memory. A within-subjects experiment was conducted in which participants listened to 24, 60-second radio advertisements that had been coded as either high- or low-imagery messages. Secondary-task reaction time was obtained during exposure to half of the advertisements, within each level of imagery. Self-reported involvement with the message was also obtained after exposure to each advertisement. Secondary-task reaction time was faster during exposure to the high-imagery advertisements. Self-reported involvement was greater for high-imagery advertisements, compared to low-imagery advertisements. Results of this study indicate that people allocate more controlled cognitive resources to encoding high-imagery radio advertisements than may be required by the message.