Not great. While there was ample food available for the wealthy and for certain skilled workers, the vast majority of the urban population was made up of unskilled day-laborers who hovered dangerously close to malnutrition. Famine was a constant threat in the city, which demanded a constant supply of quite a lot of grain which could be easily disrupted with catastrophic results--Garnsey's book on famine identifies fifteen major food crises between 104-36 BC, and every few years there were minor crises which caused grain prices to shoot up so high as to become unaffordable. Measures to impose price restrictions were generally ineffective until the Principate, and state-subsidized grain was not enough to feed a family, a necessity among the urban plebs, who generally supported rather large households. A user on this thread provided some very interesting calculations, suggesting that the state-subsidized grain was enough to provide for a single laborer (provided he was eligible to receive it) but was totally insufficient for a family. The ubiquity of families means the urban plebs had to purchase additional grain on the market, an expensive necessity even in the Principate--and, mind you, in the Principate access to the state-subsidized grain was a privilege reserved generally for a relatively well-off social level, not for the urban poor at all, who would be forced to purchase all their grain at market prices. As you can see from the linked thread, Roman wages were poor, irregular, and insufficient. A worker who could not find work for a day went without, as did his family. While outright starvation was probably mostly a threat to the old and sick, hunger and malnutrition were common in the city--moreover, the urban poor often could afford to eat little but grain, and supplementing their diet with fruit or nuts was probably relatively rare