The origin and evolution of magnetic fields (MFs) of young neutron stars (NSs) is an open question. MFs could be generated through a dynamo during the formation of NSs, or they could be a relic of a pre-supernova magnetic field. We want to test whether MFs of young NSs are the relics of their progenitors, massive OB stars. This could happen through magnetic flux conservation; the MF of massive stars is core confined, so the collapsed core might keep exactly the same magnetic flux as the whole star. Only 5-7% of massive OB stars have well-measured magnetic fields (reaching tens of kG). They can be divided into two groups: highly magnetic stars and weakly magnetic stars. NSs are also divided into normal pulsars and magnetars. We therefore assume that normal pulsars are descendants of weakly magnetic stars, while magnetars originate from highly-magnetic OB stars. To test this hypothesis, our population synthesis code takes into account some severe selection effects in the NS sample, and enables us to compare observed fractions of pulsars and magnetars with the observed fractions of weakly magnetic and highly magnetic OB stars. We also investigated independently the distribution of MFs of massive stars using the maximum likelihood technique.