Introduction: Hand-held dynamometry is a simple, standardized, relatively inexpensive way to assess decline in muscle strength in ALS patients. Comprehensive assessment of patients includes measurements of functioning and performance of upper limb activities, using standardized global assessment scales and upper limb tests of function and activity. Purpose: To assess reliability of a testing protocol for hand-held dynamometry in healthy subjects, to determine the appropriate way of reporting the outcome, to consider the usability of the upper limb muscle strength measurements with the TRICALS protocol, grip and pinch and ARAT assessment in patients with ALS. Methods: 21 healthy subjects and 30 ALS patients participated in the study. ANOVA variance analysis was performed for comparison of possible outcome choices, standard error of measurement (SEM) was reported and construct validity between ARAT, hand-held dynamometry, grip and pinch strength and ALSFRS-R. Reliability of a single rater was reported for healthy subjects using the intraclass correlation coefficients. Measurement were performed twice between 9 or 17 days. Construct validity was evaluated with Spearman correlation coefficients between all outcome measures. Results: ANOVA found no significant differences between various methods of selecting the outcome in hand-held dynamometry. Compared with the maximum measurement, the lower SEM was measured for the average of three measurements, and this outcome was used for further analyses. Reliability of a single rater in healthy subjects was good to excellent on the dominant side (ICC = 0.799–0.963) and moderate to good (ICC = 0.641–0.931) on non-dominant side. The concurrent validity between ARAT and hand-held dynamometry was good to excellent (ro = 0.765–0.913), between ALSFRS-R and hand-held dynamometry fair (ro = 0.409–0.699), between grip and pinch grip strength and ARAT moderate to good (ro = 0.535–0.804) and between ARAT and ALSFRS-R moderate to good (ro = 0.527–0.741). Discussion and conclusion: We recommend using an average of measurements as a result of hand-held dynamometry or that the reported results are a matter of agreement and should be always the same. In healthy subjects, we showed good reliability for a single rater. The calculated coefficients of correlation between the measured tests showed a fair to excellent correlation between tests in patients with ALS, which confirms that they assess similar function. Some of the muscle strength dynamometry measurements showed a floor effect in the studied group of patients. For more accurate assessment of leftover function in this group, we recommend that performance, ability and function tests (ARAT) or global function scales (ALSFRS-R), are performed. Future research should focus on the reliability of hand-held dynamometry for a single rater and between raters in patients with ALS.