Well you need to see the related question http://astronomy.stackexchange.com/questions/10106/brown-dwarfs-and-planets , because the answer to your question depends on how you define a planet.

If you demand that a "planet" has a rocky core then it seems very unlikely that a planet could form in isolation away from a parent star. The parent star is needed in order to differentiate the rocky material from the gas and allow it to condense.

On the other hand, if you wish to define a planet as simply an object below a certain mass (say the deuterium burning threshold at 13 Jupiter masses) then it seems very likely that such an object *could* form in isolation. They would be entirely gaseous, but there would be little to distinguish them from brown dwarfs at only slightly higher masses.

At present there are plenty of candidate "free-floating planetary mass" objects. 
For example see [Joergens et al. (2014)][1]; [Liu et al. (2013)][2]; [Zapatero-Osorio et al. (2000)][3]. Unless we have our understanding of the physics completely wrong, then it is likely that at least some of these are lower than 13 Jupiter masses. However, their origin remains unclear. It is *possible* they could all have formed around stars and then subsequently been ejected, but the presence of significant numbers of these objects in young star forming regions and the *lack* of $\sim$10 Jupiter-mass objects orbiting stars, suggests that there is an alternative formation scenario that can produce such objects in isolation.


  [1]: http://arxiv.org/abs/1407.7864
  [2]: http://adsabs.harvard.edu/abs/2013ApJ...777L..20L
  [3]: http://adsabs.harvard.edu/abs/2000Sci...290..103Z