And where do you get the hydrogen from? Electrolysis? A non-starter as it consumes much more electricity than is produced by the fuel cell later on. Reformation of a hydrocarbon? Yes, but where does the hydrocarbon come from? Chemical combination? Also overall an energy-intensive and potentially polluting method.
Hydrogen-powered vehicles make a net increase to pollution by reducing the efficiency already available in the process of using available fuels. Moving pollution from one place to another is not reducing it. Hydrogen is not a fuel in the sense of being an energy source, it is a carrier or storage medium like a battery. You have to manufacture hydrogen, you can't dig it up and the energy input is always greater than you get back by burning hydrogen. Furthermore, to store useful amounts of hydrogen in a vehicle requires massive fuel tanks under great pressure – not impossible but undesirable for several reasons.
The only ecologically-sound sources of hydrogen are currently produced by (wasting) hydroelectricity by (wasting) solar electricity or reformation of (wasting) biogas or perhaps from nuclear if you regard that as ecologically-sound. Reforming hydrocarbons fails to use a lot of the energy tied up in the carbon content. It is better to put hydrocarbon fuel into an efficient vehicle and use it directly, or to use a direct-reformation (of methane, for example) fuel cell in a clever arrangement.
At first glance, hydrogen would seem to have some things going for it as an alternate energy resource. Hydrogen burned in oxygen forms only water vapor. Which is a relatively benign pollutant. But when hydrogen is burned in air, more noxious oxides of nitrogen can also result. Hydrogen can directly generate electricity in a fuel cell. While replacing Carnot heat engine restrictions with a new set of efficiency limitations. The modest (5%) hydrogen injection into an otherwise conventional ICE appears to significantly improve performance. Although it is not yet clear whether net energy gains can result or how well this can be integrated with ongoing ICE improvements.
The first really big negative is that no large source of terrestrial hydrogen exists. While there a few remote wells that do produce a few percent of hydrogen, this gets normally burned off as an unexploitable waste product. Instead, hydrogen is normally produced commercially by the reformation of methane. Electrolysis is not normally a useful means of producing bulk hydrogen energy because of its staggering loss of exergy. Especially from an on-grid or pv source. Put another way, Electrolysis for bulk hydrogen energy is pretty much the same as 1:1 converting US dollars into Mexican Pesos.
A second negative is that the energy density of hydrogen is very low. The contained gravimetric density is usually lower than gasoline, while the volumetric density is up to a 3000:1 difference.
A third really big negative is that no personal vehicle practical means of storing hydrogen is known today. Compressed gas has far too little energy density, besides being a deadly terrorist bomb. Cryogenics are inefficient and expensive, besides offering only a fraction of gasoline density. And (because of a necessary boiloff) only being useful for shorter term storage. There are also frostbite and blindness safety issues. Hydrides remain expensive, low density, cumbersome, and of low lifetimes. Sadly, early enthusiasm over carbon nanotube storage has waned due to failures to replicate early spectacular claims.
Other negatives do include hydrogen having one of the widest explosive ranges known. Hydrogen flames have very low visibility, owing to emissivity mostly in the ultraviolet. The very real hydrogen safety issues get compounded by perceived "Hindenburg" lore. Hydrogen also lacks odorants or colorants and tends to rot most metals through a process called embrittlement.