The aim of this study was to determine whether the glycaemic/insulinaemic responses to hay with non-structural carbohydrate (NSC, soluble carbohydrate) of 17% (HC), 10% (MC) or 4% (LC) differs in control horses and whether these responses differ between control and horses with polysaccharide storage myopathy (PSSM). Five clinically normal control horses and seven PSSM horses, all unfit and of Quarter Horse breeding (age 9.4 ± 3.4 years, body condition score range: 4.5–6). A crossover design compared the HC and LC hay, with horses randomly assigned to hay type for 5 days, and all horses fed the MC hay during washout, after which the diets were switched. Horses were fed 1.5% BW (as fed) divided into 2 feeding per day, no grain. On morning of the fifth day of each block (seventh day for washout), horses were given 0.5% BW in hay, blood was drawn before and every 30 min for 5 h after feeding, and the rate of intake was measured. Whole blood glucose and plasma insulin were measured. The intake rate was significantly higher for HC. In control horses, the insulin area under the curve (6891.7 ± 3524.2 HC vs. 1185.4 ± 530.2 LC) was significantly higher than LC. Polysaccharide storage myopathy horses had significantly higher glycaemic and insulinaemic responses to HC vs. LC, however; the magnitude of insulin response was lower and glucose response higher in PSSM vs. control horses. Results suggest that insulin responses can differ significantly with the NSC content of hay. Feeding hay with 17% NSC produces elevations in insulin that could be detrimental for PSSM horses.