Interesting question. One distinction is the distance dependence. As is well known FRET efficiency goes as the sixth-power of the dipole separation. In the case of electron transfer the dependance on distance is exponential and the phenomenon tends to be effective at shorter distances, in the order of 10 angstroms. 

I would say that it is very-very difficult to do in practice... Depending of the system there are possibly some indirect ways to tell.

In a donor-acceptor system, electron transfer plays a critical role for the quenching. I think the critical factor is whether distance and energy level offset are enough for exciton separation.

May be the only principle difference between ET and energy transfer is that there is no requirement on the transition dipole moments of D and A for ET, while for energy transfer the dipole moment cannot be zero.

As the Forster distance gets smaller and smaller the distance dependence of the transfer looks more like an exponential as fdor PET. So the time dependent decays will look the same for both cases. I suggest you look at the D and A chemistry and ask whether PET is likely for your molecules.

The problem is that experimentally there is no way to test distance dependence accurately for a given system... Only in specially designed model systems it can be done.

As a rule of thumb, of D and A are far away - it is Förster, if close - can be anything

In case of energy transfer, the acceptor still shows emission. In case of electron transfer the fluorescence of the acceptor is quenched/not observed because the acceptor is converted into the radical anion by photo induced electron transfer. The charge transfer state can be probed with various time resolved techniques, like transient absorption.

For a more detailed description, see below.

In case of Forster energy transfer, there has to be spectral overlap between the emission of the donor and the absorption of the acceptor.

Thesis Introduction to Electron Transfer

Yes, this is a very good point! So, if we still see the acceptor fluorescence it should be FRET. However, in many cases there is no emission of the acceptor molecule even if the transfer was by Förster. There are no requirements on the emission yield of the acceptor in the Förster mechanism. 

Most things do fluoresce a tiny little bit, if you look hard enough.

We even saw the fluorescence of C60 derivatives, after energy transfer from a "PDI" dye.

See:

Article Fullerene C60–Perylene-3,4:9,10-bis(dicarboximide) Light-Har...

Leon's suggestion of an exponential dependence applies to Dexter energy transfer, not charge transfer. Even if distance dependent measurements were an option for you, charge transfer (CT) typically occurs when fluorophores are in such proximity that deviations from the sixth-power dependence of Förster theory are common due to failures of one or more of the following assumptions: point dipole approximation, weak coupling limit, dynamic average of isotropic dipoles (kappa^2=2/3), etc (for details read the work of Krueger, Vogel, Seidel, etc).

Rene is correct to suggest that if you have a fluorescent acceptor, you should look for this fluorescence to confirm energy transfer (ET). Similarly, in its absence steady-state (i.e. cw-)Photo-Induced Absorption experiments may identify absorption features from charge separated states (i.e. cation, anion). I assume the Donor & Acceptor are near enough to undergo CT. In such cases, the acceptor emission may also be partially quenched due to dye interaction - though this is often accompanied by changes to the dyes' absorption spectra or disagreement between absorption spectra and fluorescence excitation profiles (see e.g. studies concerning static quenching or hetero-dimer formation). If you see weak or negligible steady state emission from the acceptor, you should try time-resolved studies.

Time-resolved photoluminescence of the acceptor would be present, with an induction time roughly matching the donor fluorescence decay time for the case of ET. (Note that homogeneities in dye distance or orientation lead to differences in donor decay and acceptor induction times). Even acceptor dyes with small quantum yields often show some intrinsic short lived fluorescence. The acceptor fluorescence will decay more quickly if it is partially quenched due to dye interactions. Finally, transient absorption may allow you identify wether charge separated species result from CT, only excitonic features (i.e. only ground state bleaching & stimulated emission) result due to ET, or both are present. This should work even for non-Fluorescent acceptors. These states may not live long enough or have sufficient enough oscillator strengths to be seen via steady state methods.

For Forster energy transfer, taking into account the overlap integrals between donor emission spectrum and acceptor absorbance spectrum, you can calculate the Forster radius of energy transfer and the probability for the process to occur at the given average distance between the donor and the acceptor (if the distance is known) (there is one uncertainty related to the mutual orientation factor of the two dipole transitions, but for the random orientation it is 2/3). Of course, time-resolved measurements of acceptor fluorescence or S1-Sn absorbance would also help.

For electron or charge transfer, crucial is the free energy of electron transfer process determined by the redox properties of the donor and the acceptor. This process requires shorter distances between D and A (shorter than ~10 A). Furthermore, this process often leads to unexcited D and A radical pair rather than the excited state of the acceptor.

Hi

Alexander here gave some good pointers but in order to experimentally assess them you would need to perform some measurements. Following steady-state fluorescence or excitation spectra or fluorescence decays would be sensible, but you have to perform different measurements at different conditions to see some gradual change in distance. This really depends on what macromolecule you are attaching the fluorophores to. If it is a protein, you can, for instance graduall unfold or refold it by adding denaturant/buffer to an solution in which your protein is initially folded/unfolded. If it is a ssDNA, you can add increasing concentrations of salt.

Anyway, in FRET you would expect an inverse relation between donor fluorescence and acceptor fluorescence between spectra of between amplitude-weighted average lifetimes, while in PET you would expect just one dye change its spectral characteristics.

All of this is good as long as you are comparing two purely separate PET and FRET processes. In reality you can have sometimes a mixture of both. In a mixed situation, the ratio of donor fluorescence to acceptor fluorescence will be affected both by FRET but also by PET. If there was only FRET, you would see the changes in this fraction as a totally symmetric transition curve, in reference to transition midpoint. Including PET as another source of donor fluorescence reduction will yield bias from a symmetric transition curve.

Another point,more subtle and complex, but worth sharing:

FRET and PET have not only distance dependencies but are also influenced by distance fluctuations occurring at fluorescence lifetime scale (see for FRET Jacob MH, Dsouza RN, Ghosh I, Norouzy A, Schwarzlose T, Nau WM. Diffusion-enhanced Förster resonance energy transfer and the effects of external quenchers and the donor quantum yield. J Phys Chem B. 2013 117(1):185-98. & Haas E, Steinberg IZ. Intramolecular dynamics of chain molecules monitored by fluctuations in efficiency of excitation energy transfer. A theoretical study. Biophys J. 1984 46(4):429-37 ; and for quenching Attila Szabo, Klaus Schulten and Zan Schulten. First passage time approach to diffusion controlled reactions. J. Chem. Phys. 72, 4350 (1980)).

Basically, the rate of energy or electron transfer increases whenever the distance between donor and acceptor becomes smaller, therefore both FRET and PET are enhanced when intramolecular diffusion shortens the distance between them!

The difference is that with PET, you have a tight distance range at which the PET phenomenon takes place. According to the Szabo-Schulten-Schulten (SSS) theory (last reference given here) this tight distance region can be thought of as a single distance which behaves as an "absorbing" boundary (not optical absorption but the energy of the donor gets absorbed by being transferred to the acceptor). In FRET you have got many donor-acceptor distances which behave as "absorbing" boundaries, with different  "absorption" efficiencies (according the the 6th power dependence of FRET).

So, you can attach your dye molecules to a flexible macromolecule and measure FRET vs. PET in different viscosities (with Glycerol or Sucrose) and use the references above to assess the different dynamical effects of FRET vs. PET. You can use the SSS theory with multiple absorbing boundaries having multiple effieiencies and distinguish between PET and FRET.

Hi

Another way to experimentally identify a PET and FRET is by performing Transient absorption spectroscopy. Lets say you excite the donor system and simply follow the photo-bleach (PB) dynamics of the donor system. If there are no overlapping photoexcited species the PB(donor) and PB(acceptop) kinetics will be a easy way out. In case of PET, the PB of the donor remains long-lived as the excited state has lost its electron to the acceptor and the donor has not yet relaxed. On the otherhand for FRET or any energy transfer, the PB of the donor will decay quickly in the presence of acceptor as the excitation on the donor is now 'transfered' to the acceptor system.

From a spectrochemical point of view, an energetic mis-match between the donor and acceptor is important for PET where for FRET, it is important that the absorption of the acceptor should have some overlap with the PL of the acceptor as pointed out by the others before this reply.

You may look at the following pub for more details: http://www.nature.com/srep/2013/130625/srep02073/full/srep02073.html

http://www.nature.com/srep/2013/130625/srep02073/full/srep02073.html

Time-Resolved Laser spectroscopic experiment at nanosecond/picosecond level can easily distinguish the two phenomena. If electron transfer occurs one will be able to observe radical ions or radical ion pair, and if energy transfer takes place one can observe the excited state of the donor/acceptor.

If in your system possibility of both then please go through this paper I guess you may ended up with good ideas

http://pubs.acs.org/doi/abs/10.1021/jp063038s

 Hello everybody. Sorry, I know I am late but I have just found this interesting conversation. Have you observed a solvatochromic effect? The dipole moments of the ground and excited states of the donor and acceptor should change upon photooxidation processes and these changes can be more or less stabilized by the solvents. 

Following article may be useful

http://scitation.aip.org/content/aip/journal/apl/106/13/10.1063/1.4916946

In FRET, transition from acceptor is radiative while in PET it is non-radiative. PL measurement can be useful to understand this at basic level. PL decay profiles will further helps.

Read "the competition of charge and energy transfer processes in Dye-labeled seniconducting nanocrystals: calibration the spectroscopic ruler"