The cow's uterus sustains bacterial contamination and active inflammation after calving as part of the normal physiological process of uterine involution. Although the definition, incidence, and risk factors for postpartum uterine diseases (PUD) have been documented in annual calving systems with high-producing cows on total mixed ration, the literature contains little information for other production systems. The objective of this study was to quantify the incidence risk and factors associated with metritis, clinical endometritis, and cytological endometritis in high-altitude tropical dairy herds in a pasture-based system. A total of 248 cows from 5 commercial dairy herds in northern Antioquia, Colombia, were enrolled in this prospective observational longitudinal cohort study. Data collection and PUD diagnosis were performed at d 0, 3, 10, 17, 24, 31, 38, 45, and 52 (±1) after parturition. Between 60 and 70 d, cows were systematically enrolled in a synchronization protocol, and herds were visited monthly for pregnancy diagnosis until 180 ±15 d. The multivariable logistic regression model of variables associated with PUD occurrence included (a) parity, season of calving, transition diet, and body condition score at calving; (b) blood calcium, phosphorus, and magnesium concentrations, and milk urea nitrogen, β-hydroxybutyrate and fat-to-protein ratio; and (c) dystocia, retained placenta, and delayed uterine involution. The incidences of puerperal metritis, clinical metritis (MET), clinical endometritis (CE), and cytological endometritis (CYTO) were 2.8, 25, 29, and 26%, respectively. Retained placenta and dystocia were associated with MET, and MET was associated with CE. Low blood calcium concentration at calving was associated with CYTO. Overall, 51.2% of the cows suffered at least 1 category of PUD during the first 60 d postpartum (DPP). However, cows also suffered from different consecutive clinical events of PUD (MET, CE, and CYTO) during the follow-up period (0 to 180 ±15 DPP). These events could be grouped into 4 types (which were referred to as "clinical patterns") based on interactions between MET, CE, and CYTO.