Several treatments have been proposed to shorten the time to the attainment of full oral feeding (FOF) for premature infants, but there are only a few evaluation methods useful in estimating predictors of this period. We investigated whether specific items within the disorganized sucking patterns described by the Neonatal Oral-Motor Assessment Scale (NOMAS) could estimate the time to FOF in preterm infants with feeding difficulty. Preterm infants diagnosed with a disorganized sucking pattern in the NOMAS evaluation before 50 weeks of postmenstrual age were included. Video recordings of at least 2 min of oral feeding were further analyzed retrospectively by two assessors and the premature infants who exhibited disorganized sucking patterns (n = 109) were divided into three clusters (clusters 2–4). The observational items compatible with disorganization in the original NOMAS were divided into three groups: cluster 2 (disorganized: arrhythmical), cluster 3 (disorganized: arrhythmical + unable to sustain), and cluster 4 (disorganized: arrhythmical + incoordination ± unable to sustain) and further divided into incoordination-positive (cluster 4) and incoordination-negative groups (clusters 2 and 3). Premature infants in the incoordination-positive group (cluster 4, which means stress signals) showed a median transition time of 22 days (range: 4–121 days) which was longer than that in the incoordination-negative group (median 6 days; range: 1–25 days). Univariate linear regression analysis revealed that the presence of incoordination among disorganized sucking patterns (NOMAS cluster 4 vs. clusters 2 and 3), birth weight, total parenteral nutrition (TPN) duration, non-invasive positive pressure ventilation duration, the presence of moderate to severe bronchopulmonary dysplasia, pulmonary hypertension, sepsis, small for gestational age (SGA), and necrotizing enterocolitis are associated with the transition time to FOF. In a multivariate linear regression analysis, the variables revealed to be associated with the transition time were TPN duration, SGA, and the presence of stress signals (incoordination-positive group) among disorganized sucking patterns. When selecting premature infants to be treated with swallowing therapy, it is reasonable to pay more attention to the incoordination-positive group described in the NOMAS, that is, premature infants with stress signals to shorten the time to attain FOF.