The objective of this study was to investigate the effect of the pattern of concentrate-rich feeding on subacute ruminal acidosis (SARA), its severity, and the corresponding changes in VFA concentration. Eight rumen-cannulated Holstein cows were assigned to a 2 × 2 crossover design with 2 SARA challenge models and 2 experimental runs ( = 8 per treatment). Each run lasted for 40 d, consisting of a 6-d baseline, a 6-d gradual grain adaptation, and a 28-d SARA challenge period. The 2 SARA challenge models were transient (TRA) and persistent (PER) SARA. Initially, all cows were subjected to a forage-only diet (baseline) and gradually switched to 60% concentrate (DM basis). Then, cows in the PER model were continuously challenged for 28 d, whereas cows in the TRA model had a 7-d break from the SARA diet and were fed the forage-only diet after the first 7 d of SARA challenge. Thereafter, the TRA cows were rechallenged with the SARA diet. Wireless ruminal pH sensors were used to obtain ruminal pH profiles and temperature over the experimental period. For the determination of VFA, free ruminal liquid (FRL) and particle-associated ruminal liquid (PARL) were collected once for the baseline and twice (d 20 and 40 for the PER model) or 3 times (d 13, 30, and 40 for the TRA model) during SARA, each time at 0, 4, and 8 h after the morning feeding. Cows in both models experienced SARA albeit with day-to-day variation. From the start until the first 7-d SARA, cows of both models had similar pH profiles, but during the rechallenge, SARA was more severe in the TRA model than in the PER model based on lower daily mean ruminal pH (5.93 vs. 6.15; SEM 0.058) and double the amount of time at pH < 5.8 (497 vs. 278 min; SEM 68.61, < 0.05). Mean ruminal temperature was raised during SARA compared with the baseline (38.9 vs. 38.7°C; SEM 0.057, < 0.001). Concentrations of VFA increased with increasing time after feeding ( < 0.001). In general, SARA challenge (d 40 vs. the baseline), but not the challenge model, altered VFA concentrations and profile of both FRL and PARL by increasing the amounts of propionate and butyrate, whereas total VFA concentration was less affected. Proportions of VFA shifted over the duration of SARA challenge with more propionate but less acetate and butyrate proportions with advancing days of SARA challenge, leading to the values of the last SARA day being different from the earlier days ( < 0.05). In conclusion, the TRA condition led to the higher severity of SARA, but factors beyond feed intake and VFA alterations seemed to play a role.