The purpose of the study was to determine whether 12 weeks of exercise training would affect plasma glucose and plasma insulin concentrations in young and older Standardbred mares. Eighteen healthy, unfit mares representing 3 age groups were used: young (Y = 6.8 ± 0.4 years; n = 6), middleaged (MA = 15.2 ± 0.4; n = 6), and old (O = 27.0 ± 0.2; n = 6). Pre- and post-training incremental exercise tests (GXT) were performed to measure plasma glucose and insulin concentration from immediately after, until 120 min postexercise. Training consisted of exercise 3 days/week (weeks 1–8) and 4 days/week (weeks 9–12) at a submaximal intensity (∼60% of pretraining HRmax) for ∼30 min/day. Old mares had lower levels of glucose after the GXT when compared to Y and MA mares (P<0.05). There was also a time-by-age interaction (P = 0.003) in that, at 120 min post-GXT, glucose levels had not returned to pre-exercise values. Plasma glucose concentrations, in response to acute exercise, were not altered by training for any age group (P>0.05). Prior to exercise training, O mares had higher plasma insulin compared to Y and MA mares at 120 min postexercise (P<0.05), and insulin was also elevated in O mares compared to all other time points at 120 min post-GXT (P<0.05). Training resulted in increased plasma insulin concentrations at 120 min post-GXT in all age groups (P<0.05). In conclusion, age affected glucose and insulin responses to acute exercise prior to and after 12 weeks of exercise training. With regards to insulin, 12 weeks of exercise training resulted in a postexercise rebound hyperinsulinaemia, which may be related to an increased need for glycogen repletion in the muscle. These factors are important in considering the ability of an older horse to tolerate exercise.