Background: This study examined how the frequency of in-person, telephone, text-based communication, and video interactions with non-cohabiting family and friends relates to loneliness among older adults in Japan. Methods: We analyzed data from 6,786 adults aged ≥65 in the 2024 Japan COVID-19 and Society Internet Survey (JACSIS) study. Interaction frequency with non-cohabiting family and friends was categorized into none (reference), monthly, and weekly or more. Loneliness was measured using the 3-item UCLA Loneliness Scale and dichotomized at ≥4, indicating loneliness; a ≥7 cut-off was used in sensitivity analyses. Logistic regression models adjusted for sociodemographic, lifestyle, and health-related factors. Results: For family, significant associations with lower loneliness were observed only at weekly frequencies—in-person (adjusted odds ratio [aOR]=0.764, 95% confidence interval [CI] 0.664–0.879), text (aOR=0.747, 95% CI 0.659–0.846), and telephone (aOR=0.624, 95% CI 0.543–0.718), whereas for friends, significant associations were observed at monthly frequencies—in-person (aOR=0.621, 95% CI 0.549–0.701), text (aOR=0.881, 95% CI 0.778–0.997), and telephone (aOR=0.734, 95% CI 0.659–0.819). Video calls were associated with lower odds of severe loneliness (score ≥7) in sensitivity analyses—weekly calls with family (aOR=0.736, 95% CI 0.547–0.991) and monthly calls with friends (aOR=0.656, 95% CI 0.470–0.917). Conclusion Interactions with friends show associations with lower levels of loneliness at lower frequencies than family contact. While text and phone calls are broadly associated with reduced loneliness, video calls showed an association specifically with a lower prevalence of severe loneliness, underscoring the potential for tailored communication strategies.