The church was probably founded before the Norman Conquest in 1066, and may be a thousand or so years old. What we can see of the medieval fabric is mostly from the 15th century. But the history of York's ancient churches is usually complicated and St Martin's is no exception.

Sources of evidence

The most useful account of the architectural history of St Martin's prior to the bombing in 1942 is found in the Royal Commission on Historical Monuments Report on York (vol V) published in 1981. The St Martin entry is based on the work in the 1950s of Dr Eric Gee, senior investigator to the Royal Commission and an enormously experienced and respected authority. Dr Gee had a wide ranging role and limited time for research, and his report in the National Monument Record lacks the detailed supporting evidence that a study would demand today.

Documentary evidence of the earliest period of the church is absent. A reference in the Domesday Book (1086) to a church in York dedicated to St Martin is now understood to refer to St Martin cum Gregory in Micklegate. The earliest documentary reference is from around 1160-80. Dr Gee concluded from the masonry in the west of the north wall and north of the west wall of the church that it was built in the 1080s. His view was that the church began as a small rectangular stone church in the north west corner of the present site, later extended south with the addition of an aisle, in around 1200, then eastward in the late 13th century (the blocked window in the north wall dated to about 1280), then the whole extensively enlarged to its present size by the end of the 14th century with the new north aisle occupying much of the width of the original building.

Dr Gee had the considerable advantage of being able to examine the ruined church before rebuilding began in the 1960s and he was able to identify discontinuities and features of the stonework on which his conclusions were based. But like all such studies it invites further discussion. Firstly, the custom of basing a foundation date for churches solely on the earliest datable evidence is not generally accepted today.  Current thinking accepts that many churches in York were already in existence before the Norman Conquest though not necessarily built of stone. Nor is the absence of documentary evidence significant, because churches in York that were known to have been in existence in the 1080s are missing from the Domesday Book (St Olave's in Marygate is just one example). Masonry in the north wall of the church very plausibly dates from the late eleventh century but the wall consists of a hotchpotch of masonry and even ledger stones. Reuse was normal, and during the rebuilding in the 1960s pieces of decorative voussoirs of the 13th century were found within the walls – two can now be seen to the north of the present altar. But it is difficult without more work to be sure that the north wall is on the original building line. An expansion solely east and south is perhaps surprising. Dr Gee suggested that the extension east was to add a chancel and that the church may have been built on the north side of the site because of a traditional fear of burial north of the church. That might be right, and the proximity of the river (wider than today) may have prevented building to the west. But we know that land on the north side of the church was in the church's ownership in the early 14th century when it was sold off for development.

By the 15th century our knowledge of St Martin is on much firmer ground because of the wealth of documentary evidence now in the care of the Borthwick Institute for Archives at the University of York. Indeed, St Martin is outstanding in this regard. So we know that in 1411 the tower was in a poor state of repair, and that the church was substantially rebuilt in the first half of the 15th century, confirming the stylistic evidence of the surviving building.