We studied the pregnant women, who requested a '108' ambulance in two Indian states (Andhra Pradesh (AP) and Himachal Pradesh (HP)). We conducted a cross-sectional telephone survey to study the characteristics and outcomes of those who (1) were transported using '108' ambulance, (2) were sent '108' ambulance but did not use it and (3) were not assigned a '108' ambulance. We conducted interviews within 24 hours of clients' call and followed them up at 48 hours, on the 7th and 28th day. 90% of pregnant women callers in AP and only 16% in HP were from poorer socioeconomic circumstances. 22.5% of women who were not provided an ambulance in HP lived in tribal areas. A higher proportion of women who were transported using '108' reported either a high-risk condition (AP, 22%; HP, 27%) or an early complication in pregnancy (AP and HP, 16%), compared with the other groups (AP, 18% and 8%; HP, 19% and 16%). In AP, women who were sent an ambulance but did not use it had higher prevalence of obstetric emergency (9.8%) compared with the other groups (ambulance used, 7.4%; not assigned, 4.1%). One-fifth of women in AP and one-seventh in HP delivered by caesarean section. One woman who called, but was not transported by '108', died in AP. Ten stillbirths and 22 neonatal deaths were reported in AP and 17 and 16, respectively, in HP. Strategies are required to improve '108' service for tribal areas in HP. The '108' services should be improved to reduce non-use of ambulance, especially for women who report obstetric emergencies.