This study examined the short- and long-term effects of a multifactorial intervention on frailty using a randomized, crossover trial with a follow-up survey within the Hatoyama Cohort Study. Seventy-seven pre-frail or frail older adults were randomly allocated to an immediate intervention group (IIG, n=38) or a delayed intervention group (DIG, n=39). The IIG participated in the twice-weekly multifactorial intervention class consisting of resistance exercise, nutritional, and psychosocial programs. The DIG was provided no intervention for the initial 3-month period, and both groups were crossed over for the latter 3-month period. Pre-frail and frailty were determined by the Check-List 15 (CL15) score, which was validated against Fried’s frailty criteria. Intervention effects on the CL15 score and frailty status were examined for the two 3-month periods (Analysis 1). A propensity score-matched analysis was performed to balance the baseline characteristics between the participants (IIG and DIG) and non-participants of the intervention within the same cohort, and follow-up was conducted 6 and 16 months after the intervention (Analysis 2). As compared with DIG, IIG exhibited significant reductions in the CL15 score (-0.36; 95% confidence interval: -0.74, -0.03) and prevalence of frailty (-23.5%; -40.4%, -6.7%), but not in pre-frailty at 3 months, which persisted at 6 months. The DIG exhibited almost similar intervention effects in the latter 3-month period. Analysis 2 showed that CL15 scores further reduced significantly in participants who received the multifactorial intervention, with a significant group-by-time interaction (P=0.036). In summary, this multifactorial intervention was effective in improving frailty status on a short- and long-term basis.