Background: General movements assessment (GMA) is a non-invasive tool for early assessment of neonatal spontaneous movements. This study aimed to assess the effectiveness of GMA in predicting the neuromotor development outcomes of high-risk infants with neonatal respiratory distress syndrome (NRDS). Methods: The results of GMA at different phases in 80 children with NRDS expected to be born between September 2020 and July 2021 were collected, and the neuromotor development outcomes were verified by Peabody Developmental Motor Scales-2 (PDMS-2) follow-up from March 2022 to May 2022. The study used the screening method and the Chi-square test to analyze the predictive value of different phases of GMA. Results: The GMA writhing movements phase showed an accuracy of 70.00%, a sensitivity of 82.76%, a specificity of 62.75%, a positive predictive value of 55.81%, a negative predictive value of 86.49%, a Youden index of 0.46, and a positive likelihood ratio of 2.22. The GMA fidgety movements phase showed an accuracy of 95.00%, a sensitivity of 96.55%, a specificity of 94.12%, a positive predictive value of 90.32%, a negative predictive value of 97.96%, a Youden index of 0.91, and a positive likelihood ratio of 16.42. The differences between the specificity, accuracy and negative predictive values of GMA were statistically different (\chi{}^2{}_2 = 9.600, p{}_{\mathrm{2}} 0.005; \chi{}^2{}_3 = 17.316, p{}_{\mathrm{3}} 0.005; \chi{}^2{}_5 = 10.268, p{}_{\mathrm{5}} = 0.001), while no statistically significant differences were found in the comparison of sensitivity and positive predictive values (p{}_{\mathrm{1}} = 1.000; \chi{}^2{}_4 = 2.690, p{}_{\mathrm{4}} = 0.101). Conclusions: GMA has a favorable predictive value for neuromotor development outcomes in children with NRDS. Clinical Trial Registration: ChiCTR2200061223.