The Autobiographical Interview, a method for evaluating detailed memory of real-world events, reliably detects differences in episodic specificity at retrieval between young and older adults in the laboratory. Whether this age-associated reduction in episodic specificity for autobiographical event retrieval is present outside of the laboratory remains poorly understood. We used a videoconference format to administer the Autobiographical Interview to cognitively unimpaired older adults (N = 49, M = 69.5, SD = 5.94) and young adults (N = 54, M = 22.5, SD = 4.19) who were in their homes at the time of retrieval. Relative to young adults, older adults showed reduced episodic specificity in their home environment, as reflected by fewer episodic or "internal" details (t (101) = 3.23, p = 0.009) and more "external" details (i.e., semantic, language-based details) (t (101) = 3.60, p = 0.003). These findings, along with detail subtype profiles in the narratives, bolster the ecological validity of the Autobiographical Interview and add promise to the use of virtual cognitive testing to improve the accessibility, participant diversity, scalability, and ecological validity of memory research.