Aim: To determine the effectiveness of ventilations in bystander cardiopulmonary resuscitation (BCPR) and to identify the factors associated with ventilation-only BCPR.Methods: From out-of-hospital cardiac arrest (OHCA) data prospectively collected from 2005 to 2011 in Japan, we extracted data for 210,134 bystander-witnessed OHCAs with complete datasets but no prehospital involvement of physician [no BCPR, 115,733;2,093; 61,075; and conventional (compressions + ventilations) BCPR, 31,233] and determined the factors associated with 1-month neurologically favourable survival using simple and multivariable logistic regression analyses. In 91,885 patients with known BCPR durations, we determined the factors associated with ventilation-only BCPR.Results: The rate of survival in the no BCPR, ventilation-only, compression-only and conventional group was 2.8%, 3.9%, 4.5% and 5.0%, respectively. After adjustment for other factors associated with outcomes, the survival rate in the ventilation-only group was higher than that in the no BCPR group (adjusted OR; 95% CI, 1.29; 1.01-1.63), but lower than that in the compression-only (0.76; 0.59-0.96) or conventional groups (0.70; 0.55-0.89). Conventional CPR had the highest OR for survival in almost all OHCA subgroups. The adjusted OR (95% CI) for survival after dividing BCPR into ventilation and compression components were 1.19 (1.11-1.27) and 1.60 (1.51-1.69), respectively. Older guidelines, female sex, younger patient age, bystander-initiated CPR without instruction, early BCPR and short BCPR duration were associated with ventilation-only BCPR.Conclusions: Ventilation is a significant component of BCPR, but alone is less effective than compression in improving neurologically favourable survival after OHCAs.Word count: 247