I’ve often heard claims bandied about (even by Ian Morris, typically an advocate for rather low population estimates for pre-modern Near Eastern and Chinese cities) that in the 1st century BC, the population of the city of Rome reached one million. I’ve always been skeptical of them. Rome simply wasn’t that large. For it to have one million people, it would either have to have been three times larger than currently imagined, or three times more population-dense than plausible, or some combination of the two. The walled city of Constantinople (the population of which possibly never reached half a million before the Ottoman era) always seemed to me a bit larger than Old Rome. The modern city of Rome, which today contains fewer than three million people, seemed to me far more than three times larger in area than the Old City of Rome.
Based on this picture, the Aurelian walls of Rome in the late 3rd century AD looked like this. They were over 18 kilometers in length and contained an area of some 3139.33 acres. Based on the old (and highly useful) rule of 100 persons per acre in premodern urban areas, this would amount to some 313,933 persons residing within the city. This is very close to the 320,000 claimed by Augustus to have received sixty denarii from him in 5 BC. From this we can conclude that in Rome briefly preceding the death of Herod the Great, at least 100% of the population was eligible for at least some welfare benefits. We can also conclude that the great symbol of urban decline in America today, the almost entirely Black-inhabited city of Detroit, Michigan, continues to have a population greater than that of Old Rome at its classical height. Further evidence exists for early Imperial, unlike late Republican, censuses counting both women and children as Roman citizens.
The first time Rome managed to achieve a population of one million people was apparently the 1920s.