PURPOSE: In football, ice-hockey, and track and field, injuries have been predicted, and hip- and knee-strength deficits quantified using hand-held dynamometry (HHD). However, systematic bias exists when testers of different sex and strength perform the measurements. Belt-fixation of the dynamometer may resolve this. The aim of the present study was therefore to examine the inter-tester reliability concerning strength assessments of isometric hip abduction, adduction, flexion, extension and knee-flexion strength, using HHD with external belt-fixation. METHODS: Twenty-one healthy athletes (6 women), 30 (8.6) (mean (SD)) years of age, were included. Two physiotherapy students (1 female and 1 male) performed all the measurements after careful instruction and procedure training. Isometric hip abduction, adduction, flexion, extension, and knee-flexion strength were tested. The tester-order and hip-action order were randomised. RESULTS: No systematic between-tester differences (bias) were observed for any of the hip or knee actions. The intra-class correlation coefficients (ICC 2.1) ranged from 0.76 to 0.95. Furthermore, standard errors of measurement in per cent (SEM %) ranged from 5 to 11 %, and minimal detectable change in per cent (MDC %) from 14 to 29 % for the different hip and knee actions. CONCLUSION: The present study shows that isometric hip- and knee-strength measurements have acceptable inter-tester reliability at the group level, when testing strong individuals, using HHD with belt-fixation. This procedure is therefore perfectly suited for the evaluation and monitoring of strong athletes with hip, groin and hamstring injuries, some of the most common and troublesome injuries in sports. LEVELS OF EVIDENCE: Diagnostic, Level III.