The present pilot study investigated the implementation feasibility, and efficacy for reducing alcohol and drug craving, of a brief, 3-session heart rate variability biofeedback (HRV BFB) intervention added to a traditional 28-day substance abuse disorder inpatient treatment program. Forty-eight young adult men received either treatment as usual (TAU) plus three sessions of HRV BFB training over 3weeks, or TAU only. Participants receiving HRV BFB training were instructed to practice daily using a hand-held HRV BFB device. HRV BFB training was well tolerated by participants and supported by treatment staff. Men receiving TAU+HRV BFB demonstrated a greater, medium effect size reduction in alcohol and drug craving compared to those receiving TAU only, although this difference did not reach statistical significance. In addition, an interaction effect was observed in analyses that accounted for baseline craving levels, wherein heart rate variability (HRV) levels at treatment entry were predictive of changes in craving in the TAU group only. Low baseline levels of HRV were associated with increases in craving, whereas higher baseline HRV levels were associated with greater decreases in craving from start to end of treatment. In the TAU+HRV BFB group, however, there was no such association. That is, HRV BFB appeared to dissociate individual differences in baseline HRV levels from changes in craving. Given that alcohol and drug craving often precipitates relapse, HRV BFB merits further study as an adjunct treatment to ameliorate craving experienced by persons with substance use disorders.