To add some plots to @VincentFraticelli's answer, the electric field for such a ring can be written as:
$$\mathbf{E} \sim \frac{1}{(x^2 + R^2)} \cos\theta, \tag{1}\label{1}$$
where $\cos\theta$ is the angle that the edge of the ring subtends at the point $P$ where the field is being calculated. The vertical components all cancel out by symmetry.
Let's imagine the ring to be a bunch of infinitesimal "point" charges $\text{d}q$ close to each other. Let's look at one of these little charges. As you move further away from the ring, two things are happening:
First, the fraction of this point charge's field which points along $x$ and $y$ changes. The $y$ component starts off (at $x=0$) as being the only component, since at the centre of the ring all the field lines point along the $y-$axis. However, as $x$ increases, each of these infinitesimal point charges $\text{d}q$ contribute more along the $x-$axis than along the $y-$axis, because $\theta \to 0$.
Second, the overall electric field falls off as $\sim 1/x^2$, since the field of each individual point charge falls off as the inverse of the square of the distance.
Of course, by symmetry, the field along the $y-$axis always cancels out, no matter how large or small it is.
As @Vincent points out, the first term in Equation (\ref{1}) (which arises from the $1/x^2$ behaviour of the electric field) decreases as you move away from the ring. However, the second term (the cosine) actually increases, since as you move further away from the ring, $\theta \to 0$, which implies that $\cos \theta \to 1$. As I pointed out, this is because as you move away from the ring, a larger fraction of the electric field of each "bit" of charge $\text{d}q$ points in the $x$ axis. (So at infinity, all of the field of the ring is pointing along the $x-$axis, even if its magnitude is zero.)
If you plot the two functions against $x$, you see the following curves:

As you should be able to make out, very close to the ring (when $x\ll R$), it is basically the cosine term that dominates how the field changes as you change $x$. However, far away, it's the $\sim 1/x^2$ term that dominates, as you would expect.