Polymer solar cells are a promising technology for future power generation. In particular the polymer poly (3-hexylthiophene) (P3HT) has attracted widespread interest with power conversion efficiencies close to 5%. [1][2][3][4] Such devices employ the bulk heterojunction device structure, where the polymer is blended with a strong charge acceptor such as a fullerene. The processing of the blend affects its morphology and the formation of domains of P3HT-rich and fullerene-rich regions. It is at the interface between these two domains that excitons are dissociated into their constituent charges, a critical step in the operation of a solar cell. Excitons are only able to diffuse a short distance during their lifetime and therefore the size of the domains should ideally be on the order of the diffusion length, maximising the number of excitons reaching the interface and undergoing dissociation. The development of accurate measurements of the exciton diffusion length is therefore important for organic photovoltaics and the optimisation of materials, processing and device structure. To date there has been a wide range of reported values for different materials obtained by techniques such as surface quenching, [5][6][7][8] volume quenching, [9,10] microwave conductivity, [11] exciton-exciton annihilation [12,13] and photocurrent modelling of solar cells. [14] Of these the surface quenching technique is probably the most used, where the organic material is deposited onto a suitable quencher, resulting in a loss of luminescence. This loss can be quantified by comparing the emission from the quenched film with that of an identical film on a non-quenching substrate and will depend on the diffusion coefficient and the thickness of the organic film. This experiment can be performed via steady-state or time-resolved with most employing the former. However, a problem in steady-state measurements is that interference effects in the layer structures used can strongly modify the amount of light absorbed. [7] Time-resolved techniques do not require absolute measurements of the luminescence as it is only the decay of the emission from the material that is actually needed, though the initial excitation profile is influenced by optical interference. Exciton diffusion in polymers occurs on a time range from 1 ps to $1 ns, [15] thus any measurements should aim to cover this range. In this communication we will describe how time-resolved measurements of fluorescence, coupled with an appropriate quencher, enable robust measurements of the diffusion coefficient. We have applied this technique to the polymer P3HT, which despite being the most used polymer in organic photovoltaic research, has had little published on its exciton diffusion. Kroeze et al.[11] reported a value for the diffusion length of 2.6-5.3 nm from time-resolved microwave conductivity measurements, depending on whether or not excitons were reflected at the polymer/air interface. Using oxygen-induced fluorescence quenching in P3HT, Lü er et al. [9] obtained a minimum value...