Methods: A retrospective computed tomography (CT)-based analysis of the endplate cavities and fusion status following PLIFs with PEEK cages was conducted by two independent observers. The term “cavity” was used to describe the endplate changes. The vertebral endplate cavities were assessed according to the size, multiplicity, location, and presence or absence of sclerosis. Results: There were 86 fixed levels in 65 consecutive patients, with a mean age of 35.44±19.60 years. The mean follow-up was 16.5±10.1 months (range, 6–57 months). Definite fusion was seen in 56 levels (65.12%) by observer 1 versus 44 levels (51.16) by observer 2. The strength of agreement was moderate. Endplate cavities were observed in 42 levels (48.84%) by observer 1 versus 47 levels (54.65%) by observer 2, with fair agreement. The strengths of agreement for the locations, multiplicity, and size were moderate, fair, and poor, respectively. Neither age, sex, etiology, levels, nor follow-up period was significantly associated with the presence of cavities. With regard to fusions, the nonunions detected by observer 1 were significantly associated with the presence of cavities (p<0.0001). However, those detected by observer 2 were nearly significant (p=0.05). Conclusions There was a high rate of unfavorable radiographic findings in the form of endplate cavities in PLIF cases with PEEK cages. A more comprehensive classification for the assessment of fusions and endplate cavities should be formulated. We strongly recommend further CT-based studies with larger sample size and longer follow-up periods.