However, the chance of harm to the baby is still under 1%, the study of almost 65,000 births in England found.
For a second birth there was no difference in the risk to babies between home, a midwife-led unit or a doctor-led hospital unit.
Midwife-led care was in general much more likely to lead to a natural birth.
The Birthplace study is the largest carried out into the safety of different maternity settings - comparing births at home, in midwife-led units attached to hospitals, those that are stand-alone and doctor-led hospital units.
All the women followed had healthy pregnancies and began labour with no known risk factors.
It found that, overall, birth is very safe wherever it happens.
The rates of complications, including stillbirth or other problems affecting the baby, was 5.3 per 1,000 births in hospital compared with 9.3 per 1,000 home births.Prof Peter Brocklehurst, who led the research, said there were clear differences between women having their first baby and those having subsequent children.
He said: "The risk of an adverse outcome for a baby are higher for a woman planning her first baby at home than in all of the other settings, but there was no difference between the midwife and hospital obstetric units."
About 45% of women planning to have their first baby at home were transferred during labour, although this was mainly because of delays in giving birth and the need for an epidural pain-relief injection, rather than because the baby was in distress.The transfer itself was not thought to be responsible for the difference because there was no raised risk for women moved from stand-alone midwife units to hospital during labour.
There was no difference in risk when women were having their second baby, whether that was at home, in a midwife unit or a traditional hospital setting.
The rate of transfer from home to hospital was much lower too, at just 12%.
Currently, about 90% of babies are born in medically-led hospital obstetric units. And in many areas women have a limited choice of where to have their baby.Continue reading the main story
This Oxford University research raises fundamental questions about maternity care in the UK. Nine out of 10 babies are born in medically-led obstetric units. There has been a trend to centralise this into fewer and larger centres to guarantee consultant cover. Many of the decisions have taken place without definitive evidence about the safety for babies and the experience for mothers. This study provides that.
It reveals an unexplained difference in the rate of normal birth between units run by midwives and those run by doctors. The disparity on emergency Caesarean sections is particularly striking. It suggests a different culture in the way midwives and doctors see birth, with doctors concerned about risks and midwives focused on normality.
Although all politicians promise women choice about where to give birth, it is hard to see how that is meaningful. Only half of trusts in England have a midwife led unit and this research suggests it is an attractive option for the 350,000 low risk women giving birth each year.Mary Newburn, from the National Childbirth Trust (NCT), said this research should drive an an expansion in midwife-led care, either at birth centres or at home for the half of women expected to have a low-risk birth.
"It's so disappointing that, at the moment, in 50% of NHS trusts there are no midwife-led units. And only 3% of births are home births.
"We think those figures show women don't really have access to out-of-hospital options."
The research also confirms that midwife-led care is much more likely to lead to a normal birth - without any interventions, including forceps or ventouse.
That was true whether the baby was born at home or in a midwife-led unit.
The emergency Caesarean rate for the low-risk women in the study was 11% in doctor-led units compared with only 2.8% at home, and 4.4% in a midwife led unit on a hospital site.
Louise Silverton, from the Royal College of Midwives, said: "Where a woman needs an emergency Caesarean section for their first birth, they will not be regarded as low risk for the next birth, and won't have the choice of going outside a medically-led unit."
The Royal College of Obstetricians and Gynaecologists (RCOG) said it supported the expansion of midwife-led units located on hospital sites, as there can be rapid transfer if complications occur.
Although researchers say they could not find an explanation for the higher level of interventions in units run by doctors, RCOG president, Dr Tony Falconer, said: "Within an obstetric unit, care is provided in a multidisciplinary, multi-professional manner, involving midwives and specialist doctors.
"Midwifery and obstetric units both work to standard clinical guidelines and medical help is provided only when indicated."