The answer to this as it is in most questions that start with “what is the best …” is that it definitely depends.
It really depends what you’re looking for and what you’re looking to highlight. For instance, Richard Tabassi mentioned above that targeting is important. Most targeting is done using chemistry at the surface of ultra-small superparamagnetic iron-oxides (USPIO), which are not approved for human use by the FDA.
These are T2-shortening agents. It means it decreases the signal at the site of the contrast agent. There are methods to increase the signal to get positive contrast (i.e. making the image brighter at the site of the agent), such as this: Generating positive contrast from off-resonant spins with steady-state free precession magnetic resonance imaging: theory and proof-of-principle experiments – Abstract – Physics in Medicine and Biology – IOPscience
The Gadolinium-based agents are typically used more as blood-pool agents for angiography (looking at blood vessel structure) and blood-flow measurements. They are a better blood-pool agent because they have a strong effect on blood T1, resulting in bright blood in images, while USPIO effect on T1 is minimal.
Gd-based contrast agents have been discouraged / contraindicated in some cases of people with kidney conditions because of their strongly suspected side effects on the kidney, likely causing or accelerating nephrogenic systemic fibrosis (NSF): Nephrogenic systemic fibrosis: a serious late adverse reaction to gadodiamide. However, they are still widely used for many applications.
I doubt the long-term side effects of Gd or USPIO are all well-known. It’s very difficult to know the long-term effects of such agents (or of almost of anything foreign to the body). It’s easy for scientists simply to dismiss these concerns and assume something to be harmless using arguments like “it is entirely secreted through this or that system”, etc., but this is what we thought about Gd and it turned out to be wrong.
MRI has some organ specific contrast agents, and also some you can drink that don’t use Gd, but the only injectable general MRI contrast agent used in the US are chelated gadolinium (Gd) compounds.
For reasons discussed by Sherif Fahmy, these may cause nephrogenic sclerosis, but the odds are low if you have no renal disease (and they will do a blood test to check). Still, the Gd agents that are less-well chelated are more highly correlated with this problem (which is no doubt caused by free Gd+3 ion, which is toxic), so it would make sense to get the second generation macrocyclic ionic Gd agents that are better chelated. In the US, this presently means only Dotarem (gadoterate):
FDA Approves Dotarem® (gadoterate meglumine), first macrocyclic and ionic gadolinium-based contrast agent in USA
This, rather than older first-generation agents like OptiMARK, Omniscan, and Magnevist, which have all been associated with nephrogenic sclerosis. Unfortunately, you probably won’t get a choice, and certainly won’t if you’re not the payor. In an HMO you get what they have, and there are something like eight other (cheaper) agents besides Dotarem that you may, and probably will, get instead. If you pay for the MRI out of pocket, on the other hand, you can call around and find what the clinic uses, and perhaps even make a deal for what they’re going to use. They may be willing to order Dotarem for you.
I’m not a salesman for Dotarem and have nothing to do with the stuff. But if I had to pick a general Gd agent to be injected with for an affordable mri system, I would try to get this one. It just makes better sense.