In a randomised, counterbalanced, crossover design, eight men with type 1 diabetes (T1D; mean ± SD age: 27.6 ± 11.4 years) reduced insulin (INS) by 50% of their normal dose or consumed carbohydrates equivalent to 1 g of carbohydrate/kg of their body weight without the usual insulin bolus (CARBS) over two sessions, held a week apart. Each session included standardised meals, a 45-minute treadmill-walk at 7.24 km.h-1 and a six-minute walk test (6MWT). Rate of perceived exertion (RPE), blood glucose, blood ketone and blood lactate measures were taken before, during and immediately after the aerobic exercise. The distance covered in metres and the predicted VO2 max (mL⋅kg−1⋅min−1) were also calculated for the 6MWT. Participants completing the INS intervention spent more time in normoglycaemia (242 ± 135 min vs 88 ± 132 min; P < 0.01) and less time in hyperglycaemia (41 ± 95 min vs 154 ± 125 min; P = 0.01) as compared to the CARBS intervention. Mild hypoglycaemia occurred in two participants during INS and no participants during CARBS. Furthermore, there was no significant difference for blood lactate, ketone, RPE, distance covered and predicted VO2 max between interventions. Based on this pilot study, INS intervention appears to be the best approach for maintaining blood glucose levels in those with T1D during aerobic exercise, though this does need evaluating in other groups, including women, children and those with sub-optimal glycaemic control. Trial registration: ACTRN12619001397101p. Registered 09 September 2019, http://anzctr.org.au/Trial/Registration/TrialReview.aspx?id=378264