Background and purpose Since early treatment of acute stroke is associated with an improved outcome, emergency medical service (EMS) transport of stroke patients is recommended. It remains unclear, however, whether EMS transport leads to faster treatment. The impact of the transport mode on pre‐ and in‐hospital processes of care was therefore investigated. Methods The present study was based on a prospective database of 158 hospitals of the Stroke Register of Northwestern Germany, which included 162 511 stroke patients admitted between January 2010 and December 2011. Main outcome measures were the baseline characteristics associated with EMS transport and process‐of‐care indicators according to the transport mode. Results Overall, 101 850 (72.0%) patients were transported by EMS and 39 324 (28.0%) by self‐transport. The baseline characteristics showing the strongest associations with EMS use were the care situation institutional care, adjusted odds ratio (OR) 7.81; 95% CI 6.86–8.90, a disturbed level of consciousness (adjusted OR 3.00; 95% CI 2.59–3.48) and having a subarachnoid (adjusted OR 2.79; 95% CI 2.24–3.49) or intracerebral hemorrhage (adjusted OR 2.26; 95% CI 1.92–2.67). For self‐transport patients the probability of being in a higher onset‐to‐door time category was 4.36 (95% CI 4.26–4.47) and the probability of being in a higher door‐to‐imaging time category was 1.32 (95% CI 1.28–1.36). Compared with self‐transport, EMS transport was independently associated with thrombolysis (adjusted OR 1.95, 95% CI 1.77–2.15). Conclusions Patient transport with EMS was independently associated with faster hospital arrival and shorter time periods from hospital admission to brain imaging and to the frequency of thrombolysis.