Background and aims: onset 1 was a randomised trial evaluating fast-acting insulin aspart (FA) in adults with type 1 diabetes (T1D) over 52 weeks in two 26-week periods. Method: subjects were administered double-blind mealtime FA, insulin aspart (IAsp) or open-label post-meal FA, each with insulin detemir, for 26 weeks. Subjects on mealtime FA (n=381) and IAsp (n=380) continued to the additional 26-week period assessing long-term safety and efficacy. Results: after 52 weeks, HbA1c change from baseline (-0.08% [FA] vs. +0.01% [IAsp]) showed significant estimated treatment difference (ETD) [95% confidence interval (CI)] favouring FA (-0.10% [-0.19;-0.00]). Change from baseline in 1-h postprandial plasma glucose (PPG) increment after a standardised meal test significantly favoured FA (ETD: -0.91 mmol/L [-1.40;-0.43]); a similar trend was seen in change from baseline in 2-h PPG increment (ETD: -0.42 mmol/L [-1.11;0.27]). Mean 7-9-7-point self-measured plasma glucose profiles significantly favoured FA (ETD: -0.23 mmol/L [-0.46;-0.00]). Median total insulin dose was 0.77 U/kg (FA) vs. 0.83 U/kg (IAsp). No difference was observed for body-weight change (ETD: 0.13 kg [-0.38;0.65]). After 52 weeks, adverse events were similar between FA and IAsp, and as expected for IAsp. Severe or blood glucose-confirmed hypoglycaemia rates (plasma glucose <3.1 mmol/L) were similar with FA versus IAsp (estimated ratio: 1.01 [0.88;1.15]).