In order to determine whether small-sided game (SSG) locomotor performance can serve as a fitness indicator, we (1) compared 6-a-side (6v6) SSG-intensity of players varying in fitness and skill, (2) examined the relationship of the 6v6-SSG and Yo-Yo IR2 and (3) assessed the reliability of the 6v6-SSG. Thirty-three professional senior, 30 professional youth, 62 amateur and 16 professional woman football players performed 4 × 7 min 6v6-SSGs recorded by a Local Position Measurement system. A substantial subgroup (N = 113) also performed the Yo-Yo IR2. Forty-seven amateur players performed two or three 6v6-SSGs. No differences in 6v6-SSG time-motion variables were found between professional senior and professional youth players. Amateurs showed lower values than professional seniors on almost all time-motion variables (ES = 0.59-1.19). Women displayed lower high-intensity time-motion variables than all other subgroups. Total distance run during 6v6-SSG was only moderately related to Yo-Yo IR2 distance (r = 0.45), but estimated metabolic power, high speed (>14.4 km · h(-1)), high acceleration (>2 m · s(-2)), high power (>20 W · kg(-1)) and very high (35 W · kg(-1)) power showed higher correlations (r = 0.59-0.70) with Yo-Yo IR2 distance. Intraclass correlation coefficient values were higher for total distance (0.84) than other time-motion variables (0.74‒0.78). Although total distance and metabolic power during 6v6-SSG showed good reproducibility (coefficient of variation (CV) < 5%), CV was higher (8-14%) for all high-intensity time-motion variables. It was therefore concluded that standardised SSG locomotor performance cannot serve used as a valid and reliable fitness indicator for individual players.