We aimed to (a) evaluate the agreement between ultra-short-term and criterion resting heart rate variability (HRV) measures in military trainees, and (b) compare associations between HRV recording lengths and body composition. HRV recordings were performed for 10 min in 27 military male students. Mean RR interval, the root-mean square of successive differences (RMSSD), RMSSD:RR interval ratio, standard deviation of normal-to-normal RR intervals (SDNN), and SDNN:RR interval ratio were determined from the last 5 min of the 10-min recording and considered the criterion. Parameters were also recorded in successive 1-min epochs from the 5-min stabilization period. No differences were observed between criterion values and any of the 1-min epochs (p > 0.05). Effect sizes ranged from −0.36–0.35. Intra-class correlations ranged from 0.83–0.99. Limits of agreement ranged from 38.3–78.4 ms for RR interval, 18.8–30.0 ms for RMSSD, 1.9–3.1 for RMSSD:RR, 24.1–31.4 ms for SDNN, and 2.5–3.0 for SDNN:RR. Body fat% was associated (p < 0.05) with all HRV parameters at varying time segments. A 1-min HRV recording preceded by a 1-min stabilization period seems to be a suitable alternative to criterion measures. Ultra-short procedures may facilitate routine HRV tracking in tactical populations for status-monitoring purposes.