A Danish study is now saying that women who drink moderate amounts of coffee while pregnant should be reassured they are not increasing risks to their baby.
Earlier work found high caffeine intake could increase the risk of premature birth and having a small baby.
But the British Medical Journal research found no difference between women who drank moderate amounts of caffeinated and decaffeinated coffee.
UK pregnant women are advised to drink no more than 300mg of coffee per day.
This equates to three cups of instant coffee per day, six cups of tea or eight 40mg cans of cola or eight 40 mg bars of chocolate.
The Food Standards Agency advises having a bar of plain chocolate, drinking three cups of tea, a can of cola and a cup of instant coffee in a day would meet the 300mg limit.
The researchers from the University of Aarhus monitored 1,207 healthy pregnant women who had what was classed as a high caffeine intake of more than three cups of coffee a day.
All were less than 20 weeks pregnant when they joined the study.
Half the women were given normal coffee and half decaffeinated coffee so researchers could evaluate who was most at risk of giving birth early or of having a small baby.
Neither the women, nor the researchers who gave out the coffee knew which sort was being given.
Each woman was also interviewed to check if they were drinking other drinks containing caffeine, such as tea and cola.
The length of pregnancies and the baby's birthweight was also recorded.
Age, pre-pregnancy weight and smoking habits were also taken into consideration.
It was found that there was no real difference in either the length of pregnancy or birth weight between the two groups.
Women who had the decaffeinated coffee had a caffeine intake which was 182g lower than those who drank the caffeinated form.
Among women who drank the caffeinated coffee, 4.2% of babies were born prematurely and 4.5% were small for their gestational age.
In the group drinking decaffeinated coffee, 5.2% of babies were born prematurely and 4.7% were underweight.
There was virtually no difference between the average birthweights in each group.
The researchers said decreasing caffeine intake during the later stages of pregnancy has no overall effect on birth weight and length of pregnancy.
However, they said any effect in the first half of pregnancy would not have been detected by their study.