Background: The Interpersonal Psychological theory of suicide (IPTS) remains controversial due to mostly heterogeneous study findings. Previous studies on the IPTS are mainly based on cross-sectional data and group-based models. This assumes a one-size-fits-all model and masks potential between-person variability in the degree to which the predictors thwarted belongingness (TB), perceived burdensomeness (PB), and hopelessness are associated with suicidal ideation. This study aimed to demonstrate that there is substantial between-person variability in the way these predictors are associated with suicidal ideation. We further aimed to identify clusters of patients for which the IPTS is more or less applicable and to determine whether clusters differ in overall suicidal ideation, depression, and positive affect. Method: Ecological momentary assessment data of N=74 psychiatric inpatients (M = 37.6, SD = 14.3; range: 18-85; 71.6% female) was analyzed with lagged multilevel VAR models and idiographic associations were modeled. K-means clustering was used to identify clusters of patients and clusters were statistically compared with regard to specific patient characteristics. Und Results: We found substantial between-person variability in the degree to which TB, PB, and hopelessness predicted suicidal ideation. Cluster analyses revealed four clusters, including a large cluster of people for which none of the predictors was predictive for suicidal ideation (n=36). Cluster analyses revealed that different predictors were relevant for different groups. Clusters for which the IPTS was more applicable were characterized by comparably high suicidal ideation, depression, and low positive affect. Discussion: Our findings suggest that a one-size-fits-all model for suicide prediction does not reflect between-person differences in processes leading to suicidal ideation. A promising avenue for future research might be to use idiographic approaches and identify subgroups of people to determine individual risk factors and personalize prediction and treatment.