Objectives: This study investigated the inter-rater reliability of three structural end range lumbar segmental instability tests with the highest positive likelihood ratio ( z LR) against flexion-extension radiographs, and three functional mid-range clinical tests that predict the success of lumbar stabilisation exercises in patients with recurrent or chronic low-back pain (R/CLBP). The study also investigated the reliability of lumbar segmental instability, subclassification as: functional, structural and combined instability. Method: Forty adults with R/CLBP (30 men and 10 women), aged 21-71 years, underwent repeated measurements of specific clinical tests for structural or functional lumbar segmental instability. Results: All functional-instability tests: the prone instability test (PIT), the aberrant motion test and the average passive straight-leg raise (PSLRw91u u) test showed a high percentage agreement (90, 97?5 and 95%, respectively) and a high kappa coefficient (0?71, 0?79 and 0?77, respectively). In addition, two structural tests: the lumbar flexion range of motion (ROM) w53u u and the passive lumbar extension test (PLET) showed a high percentage agreement (82 and 73%, respectively), and a moderate kappa coefficient (0?48 and 0?46, respectively). The lack of hypomobility with the posteroanterior (PA) glide test was found to be unreliable (agreement525%; k5{0?02). Locating the pain-provoking segment, as the first portion of PIT, was found to be moderately reliable (k50?41). The subclassification categories of lumbar segmental instability (functional, structural and combined) were found to be significantly reliable (PABAK) 0?90, 0?70 and 0?95, respectively). Discussion: All investigated tests (except the lack of hypomobility with the PA glide test), in addition to subclassifying the categories of lumbar segmental instability, were significantly reliable in the assessment of lumbar instability.