I disagree with the posters here, and most scholars of Caesar are in agreement with me and our texts. Caesar did not move his troops at 90 miles a day regularly, only on a few occasions to steal a march on his enemies. Still, Caesar did move very quickly through Gaul at all times, but his extremely rapid advances were only done once or twice to maneuver around enemy formations--Caesar was quite well known for appearing in the rear of an enemy army literally overnight. Both posters here correctly mention that Caesar would often march with no baggage and often resorted to forced marches, but they neglect Caesar's use of night marches, which was a key feature of his strategy of movement. While marching to Gergovia, for example, Caesar split his army in two, leaving behind some troops as decoys, and marched through the night without stopping to come around behind Vercingetorix. Against the Helvetii Caesar marched throughout the night to beat them to the Saone. The claim is made by another poster here that Suetonius claims that Caesar marched nearly 100 miles, not his army. That's simply untrue. Suetonius 57 mixes both Caesars personal movement and the movement in his army together in a single passage, without differentiating between the two:

In agmine nonnumquam equo, saepius pedibus anteibat, capite detecto, seu sol seu imber esset; longissimas vias incredibili celeritate confecit, expeditus, meritoria raeda, centena passuum milia in singulos dies; si flumina morarentur, nando traiciens vel innixus inflatis utribus, ut persaepe nuntios de se praevenerit.

The important parts of the passage are "in agmine...anteibat" and "meritoria raeda...centena passum milia in singulos dies." The first excerpt clearly shows that Suetonius sees Caesar's movement as the same as that of his army. Agmen is often translated as "march," because it literally means a "flow," or "something that is driven," but as a transferred epithet in military accounts almost always means "army," specifically an army on the march. Anteibat literally means "go before," and I could see why there might be confusion here, as the sense is not clear in English--does it mean Caesar marched ahead of his army, moving faster than it, or does it mean that he marched at the head of the army, leading it? In Latin it's much clearer--anteeo means "to lead," or "to take the lead." The only unclear part of the passage is meritoria raeda, "by hired carriage." It seems here that Suetonius is mixing Caesar's personal movement (probably his breakneck race back to Gaul at the beginning of Vercingetorix's revolt) with that of his army--the rest of the passage quite clearly describes military maneuvers, not the actions of a single individual (troops swimming the river, or floating across it on skins--this also seems to be referring to Caesar's night march at Gergovia, as the point of it was to cross the river behind Vercingetorix's back).

In any case, the argument that that is the only reference to a 90-mile march is patently false. Caesar made the march to Lake Geneva in 8 days! A quick look at geography will show this to be a rate of march between 80-100 miles a day, a speed only possibly because of Caesar's insistence on marching through the night. Were his troops tired? Goddamn right they were, and Caesar was able to delay the Helvetii for two weeks with diplomatic talk, giving his men the opportunity to rest--the whole point was to get to the Rhone before them, allowing him to destroy the bridges and force the Helvetii to stop and talk. After that two-week lull Caesar marched (again at night) to the Saone and cut a portion of the Helvetii off there, appearing overnight on a hill behind them.

Nor is such a speed of march impossible--Napoleon marched a comparable speed through Italy, slowed down only during the beginning of his campaign by the Alps. The two posters here note that if you take a normal infantry march as 3.5 mph, then 90 miles per day is pure fantasy. This is true, provided that that estimate is the speed of the march and the army is only marching during daylight. As has been shown, the duration of Caesar's marches frequently lasted into the night (several times, such as at his march to the Saone, Caesar specifically says he marched through the night "without pause"). 3.5 mph is also a pretty pedestrian march, hardly a rapid advance--if we round it off to 4 mph (slightly faster) that gives us about 22 hours in which to complete a march of 90 miles. This a decently good estimate of a standard Roman march--Vegetius says that the normal speed of march is 20 Roman miles in 5 summer hours, about 4 mph. Vegetius mentions a quick march as well, of 24 Roman miles in 5 summer hours, which is closer to around 4.5 mph. Vegetius also mentions that armies moved at a run as well--according to Vegetius anything over the quick march is a run, and no precise speed can be attributed to it. That marches were sometimes made at a run is obvious--Vegetius says that recruits should be trained extensively at running in step, so that they can not only charge the enemy at a good pace, but also so that they can march to positions faster than an enemy army, conduct reconnaissance more effectively, and pursue a routing army. I believe that Goldsworthy has also suggested that the Roman army might have employed a sort of mixed march, such as that that the British army (and the French, and just about every army of the period) employed during the Napoleonic Wars, marching for several paces, then running for several. This is certainly not implausible, though we have no direct evidence for it.

In short, Caesar's rate of march is hardly the stuff of fantasy. Difficult, surely, but not impossible. That it was used only rarely is quite clear, as was the fact that it was physically extremely exhausting on the troops--Caesar had to wait two weeks before his troops were in shape to fight again at Geneva, and for some of that period he was reorganizing units that had fallen out during the march. We should not take this to be Caesar's regular marching pace--indeed, even the briefest look at de Bello Gallico would show this to be untrue, though Caesar's marches were always rather quick, since he marched for longer than most armies. Instead, we should consider it to be a tactical decision to outflank and outmaneuver an enemy. Gergovia is perhaps the best example, and Caesar's march to Geneva is quite an outlier, being the only occasion when he extended his rapid march for such a long period of time.

Do you have the page number for Goldsworthy?

That's pretty high, even if you had mounted troops/cavalry that is a 14ish hour march which would be pretty taxing for horses and man. And that's assuming they leave their trains behind.

The Romans could move very quickly, the men were used to carrying several days worth of supplies that would allow them to move without a baggage train - greatly increasing their mobility. But if you have an average marching speed of 3.5mph then 90 mi is pretty fantastical...

Ignore this post, just leaving it up for reference. Look instead to XenophonTheAthenian post.

Although Ceasar did move his troops very quickly, 90 miles in a day is false. One of the longest recored marches was around 40 miles I think (Can't find my source). You may be drawing that 90 mile thought from Suetonius "Ceasar" P 57 where he says Ceasar marches 100 miles in a day. Ceasar not his army. He was in fact a believer in forced/loaded marches as seen in his campaigns mainly in gaul. He was able to do this because he was so charismatic and loved by his troops. He would often sever his baggage train in order to march faster which is a very risky move that can result in the cutting off of the armies. It almost always worked out for him but more often than not for other commanders it results in disaster. Edit: grammar

I have never read any estimate higher than 30 miles a day. Just Sayin'