Memory performance in everyday life (EDL) and its change by aging is still unclear, since laboratory results are often not transferable to real life situations. Therefore, a naturalistic memory task was designed to investigate memory performance in older adults in a representative task design. The memory performance of 23 older (70.28 ±4.65 years) and 20 younger adults (24.89 ±3.16 years) was assessed by using four established tasks of Short-Term and Working Memory (Digit Simple Span, Digit Complex Span, Grid Simple Span, and Grid Complex Span), which differed in complexity and domain (verbal vs. visual-spatial). To simulate an EDL context, a “Supermarket” was constructed within the laboratory. The results showed that memory performance presents alterations in older adults. This was especially true for the “Supermarket” task, in which the younger adults showed benefits in the common environment as opposed to older adults. A factor analysis showed that the four memory laboratory tasks measured different memory processes as compared to the EDL task. Our findings suggest that memory performance in EDL is a different construct than when it is measured in the laboratory and that memory alterations in older adults are especially pronounced in EDL tasks. Following these findings, we recommend an EDL approach to measure memory performance further to the well-established laboratory approaches.