Multimorbidity has become a global public health concern, yet cross-national comparisons remain limited, especially in longitudinal settings. This study investigates the longitudinal patterns and transitions of multimorbidity status of people over age 50 in 34 countries. Utilizing comparable health indicators across countries, we examine chronic health conditions (hypertension and diabetes), cognitive function, physical ability, and self-report of general health. Using latent transition analysis, we identify a pattern of multimorbidity and classify it into three classes: mild, moderate, and severe multimorbidity. Mild multimorbidity is characterized by a lower prevalence of three morbidities out of five, while severe multimorbidity is characterized by a higher prevalence across all health conditions. Moderate multimorbidity falls between these two extremes. Our findings reveal substantial variation in these classes across countries, with diabetes and hypertension emerging as the predominant condition among older adults with severe and moderate multimorbidity, respectively. Over time, both severe and moderate multimorbidity tend to increase, with similar transition probabilities from mild to more severe categories across countries. Covariate analysis indicates that men and low-educated individuals are more likely to experience severe multimorbidity. These results underscore the importance of understanding multimorbidity patterns and dynamics for effective public health planning and healthcare services.