Purpose: The aim of this study was to evaluate the reproducibility of internal and external load parameters during recreational small-sided football games. Methods: Ten healthy untrained young adult males (age: 20.2 ± 1.9 yr, body mass: 69.2 ± 6.3 kg, height: 175.4 ± 5.9 cm, body fat: 19.7 ± 5.2%) performed two 2 × 20-min sessions of four versus four plus goalkeeper small-sided games (SSG) 1 week apart on a standard, outdoor, 40 × 20-m artificial grass pitch. Twelve external (total distance, peak speed, player load, work rate and distance covered at 0-2, 2-5, 5-7, 7-9, 9-13, 13-16, 16-20 and >20 km/h) and seven internal load parameters (heart rate and time spent in different heart rate zones [<70%, 71-80%, 81-90%, 91-95%, 96-100%, 91-100%]) were measured. Reproducibility was reported as intraclass coefficient correlation (ICC), the coefficient of variation (CV), and the typical error of measurements (TE). Results: No statistical differences (p> 0.05) between sessions were found in any measures. Minimal test-retest variability was noted for mean and peak heart rate (HR[subscript peak]) relative to HR[subscript peak] with CV values of 3.4% and 2.6%, respectively. Acceptable variability (CV<10%) was demonstrated for total distance covered, distance covered at 2-5 km/h, and peak speed. Distance covered in different speed zones (CV = 15.7-47.6%) and percentage of time in each HR zone showed large-to-very large variability (CV = 36.2-128.4%). Mean heart rate (HR[subscript mean]), HR[subscript peak], distance covered at 5-7, 13-16 and >20 km/h, and percentage of time above 95%HR[subscript peak] were the most reliable variables (ICC = 0.74-0.79), followed by total distance covered, peak speed, and percentage of time at 80-90% HR[subscript peak] (ICC = 0.39-0.67). The lowest reliability was observed for distance covered in the moderate speed zones 7-9 km/h (ICC = 0.12) and 9-13 km/h (ICC = -0.09), and percentage of time at 70-80% HR[subscript peak] (ICC = -0.01). Conclusions: Small-sided games can be used when planning training-induced exercise responses in relation to total distance covered, peak speed, and mean heart rate. This evidence further supports the use of SSG when organizing recreational football training, in young adult males, with the purpose of improving health profile due to the high reproducibility of HR[subscript mean] and total distance covered.