BACKGROUND Paraspinal muscle strength and fatigue are considered important in low back pain (LBP) prevention and rehabilitation. High reliability of paraspinal strength and electromyographic (EMG)-fatigue parameters has not been universally reported. Moreover, the discriminative validity of these parameters requires further exploration, under the threat of potentially poor reliability of the methods examined. AIM To investigate the reliability and discriminative validity of paraspinal strength and EMG-related fatigue in subjects with recurrent LBP and healthy participants. METHODS Test-retest measurements were performed in 26 healthy and 66 LBP volunteers, for reliability. Paraspinal isometric maximal and mean strength were determined with a maximum voluntary isometric contraction (MVIC) protocol, performed in a custom-made device. For the fatigue test, participants performed a 60% MIVC level continuous isometric contraction of the paraspinals, in conjunction with EMG analysis from 4 muscle sites of the lumbar spine. Initial median frequency (IMF), the median frequency slope (MFslope), as well as the root mean square (RMS) slope EMG parameters were used as fatigue measures. Data were analysed with repeated measures ANOVA for test-retest differences. For reliability, the intraclass correlation coefficient (ICC3,1), standard error of the measurement (SEM) and the smallest detectable difference (SDD) were reported. Group-related differences for fatigue measures were analysed with a Multivariate Analysis of Covariance, with age, weight and strength as covariates. RESULTS Isometric strength presented statistically significant between-day differences (P < 0.01), however these did not exceed 10% (healthy: 7.2%/LBP-patients: 9.7%) and ICC reliability values were excellent, yet test-retest error was increased for the patient group (healthy: ICC3,1: 0.92-0.96, SEM: 5.72-5.94 Hz, SDD: 18.51%-18.57%/LBP-patients: ICC3,1: 0.91-0.96, SEM: 6.49-6.96, SDD: 30.75%-31.61%). For the frequency data, IMF reliability was excellent (healthy: ICC3,1: 0.91-0.94, SEM: 3.45-7.27 Hz, SDD: 9.56%-20.14%/patients: ICC3,1: 0.90-0.94, SEM: 6.41-7.59 Hz, SDD: 17.75%-21.02%) and of MF raw and normalised slopes was good (healthy: ICC3,1: 0.78-0.82, SEM: 4.93-6.02 Hz, SDD: 13.66-16.67%/LBP-patients: ICC3,1: 0.83-0.85, SEM: 6.75-7.47 Hz, SDD: 18.69%-20.69%). However, the reliability for RMS data presented unacceptably high SDD values and were not considered further. For discriminative validity, less MVIC and less steep MFslopes were registered for the patient group (P < 0.01). CONCLUSION Reliability and discriminative ability of paraspinal strength and EMG-related frequency parameters were demonstrated in healthy participants and patients with LBP.