Can a half life be given in electron volts? I'm using this link to search for particular energies in which gammas may be emitted (for nuclide identification on a gamma spectrum).
If on the above link you go down to the "γ condition #1" line, and put the energy between 2241 an 2243, and click search, you get the list of Nuclides which emit gammas between these energies. The second in the list (24Mg), in the half life column, has a halflife given in keV.

This isn't the only time this occurs, this is just one example. To my mind there's no way that you can have a half life given in keV. 
Can you, in fact, give a half life in keV (or any unit of energy)? How would that make sense?
 A: I think I found your answer in the glossary section (slightly edited):

$T_{1/2}$ [the half-life] is related to (...) the width $\Gamma$ by
  $$ T_{1/2} = \frac{\hbar \ln 2}{\Gamma}$$
  where $\hbar$ is the reduced Planck's constant. Please note that when the half-life of a given nuclear level is smaller than $10^{-15}$ seconds, it is customary to list the width of the level instead.

$\Gamma$ has units of energy, so plugging in $2.5\, \mathrm{keV}$ I get a half life of $1.82\times 10^{-19}\,\mathrm{s}$ or $1.82\times 10^{-4}\,\mathrm{fs}$.
What "width" refers to is the width of the energy curve: when measuring decays you can plot the probability of getting a particle with a specific energy as a function of said energy. This plot is bell shaped; it is centered around the average energy you get, and its width at half-maximum is called $\Gamma$. Measuring $\Gamma$ is usually how these very short half lives are determined.
A: It is in natural units that time and space may be described in energy units.

