This article presents the results of a recurrence analysis conducted on the milling of GFRP material. The novelty lies in applying variable milling parameters to assess the detectability of defects. It appears that cutting parameters play a significant role in detecting defects. The research aimed to determine the sensitivity of recurrence analysis in detecting defects of various sizes in composites machined with different milling parameters. Furthermore, the study identified the most appropriate quantifications for defects detection. The values of some quantifications derived from the cutting force showed significant changes compared to the indicators obtained during machining without defects. It was demonstrated that recurrence quantification analysis could identify defects as small as 0.25 mm, while recurrence diagrams alone could detect defect sizes of 0.75 mm or larger. The most effective indicators were RR, T1, and T2. Additionally, it was demonstrated that certain indicators depend on cutting parameters, complicating their use for defect detection.