Background: Suicide behavior represents a major public health problem for the older population. Within the continuum of suicidal behavior, suicidal ideation may lead to a suicide attempts/death. Risk factors for developing suicidal ideation include mobility limitations, lack of social participation and loneliness. However, there is a need for longitudinal studies to examine these relationships over time. Method: 50423 older people from three waves of the SHARE project formed the sample (60 years in the first wave; M f SD = 71.49 f 8.15; 55% female). Results: A series of nested Cross-Lagged Panel Models (CLPM) of suicidal ideation, mobility limitations, social participation and loneliness were tested. The best fitting model was that with equal autoregressive and cross-lagged effects across waves (chi 2 chi 2 = 1220.56, CFI = .982, RMSEA = .028, SRMR = .024). The autoregressive effects showed high stability across waves. The cross-lagged effects between suicidal ideation and mobility limitations were strong, while the cross- lagged effects between suicidal ideation and social participation were comparatively smaller. In the case of loneliness, statistical significance was not achieved. Conclusions: These findings highlight the importance of promoting mobility programs and social activities to prevent suicidal ideation among older adults.