Poor sleep is a long-term public health issue that has become increasingly prevalent among socially isolated older adults. However, research on the mechanisms explaining the link between social isolation and impaired sleep (IS) remains limited, particularly in low- and middle-income countries. This study explored the serial mediating effects of loneliness and mental distress on the association of social isolation with IS among Ghanaian older adults. We analyzed data from 1201 adults aged ≥ 50 from Ghana's AgeHeaPsyWel–HeaSeeB study (mean age = 66.14, SD = 11.85, age range = 50–111; women = 63.28%). Measures included the UCLA 3-item Loneliness Scale, modified Berkman–Syme Social Network Index, Sleep Quality Scale, and Mental Distress Questionnaire. We used bootstrapping techniques from Hayes’ PROCESS macro program to estimate the hypothesized serial mediation. Social isolation was significantly associated with IS (β = 0.242, p < 0.001). Crucially, social isolation indirectly predicted IS via three significant mediating pathways. Loneliness accounted for 17.6% (β = 0.054, CI = 0.096, 0.016), mental distress accounted for 6.5% (β = 0.020, 95% CI = 0.004, 0.040), and loneliness and mental distress accounted for 32.2% (β = 0.099, 95% CI = 0.065, 0.138) of the overall effect. The total mediating effect was 56.4%. These findings suggest that the social isolation-sleep link is respectively and serially explained by loneliness and mental distress. Social integrative interventions for sleep quality in old age should target mental and emotional well-being.