Background Within a law enforcement agency (LEA), custody assistants (CAs) are responsible for upholding proper safety and security inside correctional facilities. However, unlike other law enforcement positions, CAs may not be subjected to fitness testing prior to matriculation. If there are differences in fitness between recruits across different classes, this could influence training adaptations following academy. Purpose The purpose of this study was to investigate the physical fitness of CAs across three different academy classes. Methods A retrospective examination of performance data was conducted on 108 CAs from three classes (Class 1: males=29, females=11; Class 2: males=22, females=16; Class 3: males=18, females=12). The fitness tests encompassed: number of push-ups and sit-ups in 60 seconds; 201-meter (m) and 2.4-kilometer (km) run times; and estimated maximal aerobic capacity (VO2max) derived from the 2.4-km run. To compare males and females from the classes (the sexes were analyzed separately), a one-way analysis of variancewith Bonferroni post hoc was utilized (p<0.05). Effect sizes (d) were also calculated. Results Class 2 males executed significantly more sit-ups than those from Class 3. There were moderate effects for the greater number of push-ups completed by Class 2 compared to Class 3, and the faster 201-m run for Class 3 compared to Class 2 (d=0.60-1.00). There were no significant between-class differences for the females, but moderate effects for the greater sit-ups and estimated VO2max for Class 1 compared to Class 3 (d=0.64-0.76). Conclusion Even without physical testing prior to academy, the CA classes from this LEA seemed to be relatively similar in fitness. However, physical training instructors should acknowledge that there may be select variances between certain classes (e.g. abdominal strength measured by sit-ups; anaerobic endurance measured by the 201-m run). Instructors should utilize appropriate assessments to characterize fitness of their recruits, and where possible, tailor training accordingly.