Objective: To evaluate textural parameters of ultrasound images of reactive and neoplastic cervical lymph nodes, investigating differences between the groups and the potential of texture analysis to characterize these differences. Methods: We analyzed 36 B-mode (2D) cervical ultrasound images, corresponding to 16 reactive and 20 neoplastic lymph nodes, from examinations performed between 2010 and 2024 at a diagnostic imaging clinic in Parnaíba, Piauí, Brazil. Reactive lymph nodes were classified based on clinical course and response to treatment, while neoplastic lymph nodes were confirmed histopathologically. The images were analyzed using MaZda, with parameters based on the gray-level co-occurrence matrix (GLCM), considering different variables, directions, and distances between pixels. The Mann–Whitney test was used, with a significance level of 5%. Results: Four variables were statistically significant: Sum of Squares (SumOfSqs) in the direction S(1,-1), Entropy Difference (DifEntrp) S(2.0), Sum of Means (SumAverg) S(4.0), and Sum of Entropy (SumEntrp) S(0.4). Greater dispersion of SumOfSqs, DifEntrp, and SumEntrp was observed in reactive lymph nodes, while SumAverg was elevated in neoplastic lymph nodes. Conclusion: Textural analysis identified significant differences between the groups, suggesting its potential as a complementary quantitative approach to conventional ultrasound evaluation. Future studies may validate these findings and expand their clinical application.