Choroidal melanoma is the most frequent primary intra-ocular tumour. Although being a malignancy of the choroidal melanocytes, it affects the retinal pigment epithelium (RPE). The resulting epitheliopathy appears as an area of atrophy, drusen, lipofuscin accumulation and localized detachment over the RPE associated with phagocytic activity that changes fundus colouration (Lavinsky et al. 2007). Lipofuscin from degraded lysosomal function accumulates in RPE cells and in macrophages of malignant choroidal tumours and generally appears as orange pigmentation over the lesion. Colour may vary from orange to brown or red brown depending on whether the tumour is melanocytic or amelanocytic. Orange pigment has also been considered a sign of malignancy. Fundus autofluorescence (FAF) photography is based on the stimulated emission of light from naturally occurring fluorophores, the most significant being lipofuscin. The recently developed Heidelberg Spectralis HRA+OCTdevice (Heidelberg Engineering, Heidelberg, Germany) enables a better assessment of autofluorescence than earlier devices, and, in addition, it evaluates the spatial distribution of lipofuscin-mediated fundus autofluorescence in vivo (Singh et al. 2010). We evaluated FAF in 100 choroidal nevi and in 65 choroidal melanomas. The mean best-corrected visual acuity (BCVA) for eyes with choroidal nevi was 20/14 ± 1.14 Snellen with a lesion thickness of 1.75 ± 0.23 μm (mean ± standard deviation); the mean BCVA of eyes with choroidal melanomas was 20/8 ± 2.61 with a mean tumour thickness of 3.67 ± 1.27 μm; all patients were of Caucasian origin. In 40 eyes with choroidal nevi, FAF showed a normal pattern of background fundus fluorescence with no corresponding areas of hyperfluorescence or hypofluorescence over the nevi (Fig. 1). Sixty eyes with choroidal nevi presented hypoautofluorescence associated with chronic RPE degenerative features and RPE atrophy. In 26 choroidal melanomas, fundus photography showed orange pigmentation, and FAF showed a plaque-like hyperautofluorescence throughout the surface of the lesion, which corresponded to the areas of increased pigment and fibrous metaplasia. Moreover, macrophages invade the highly pigmented areas in the attempt to phagocytize the accumulated material (Fig. 1). The lipofuscin overlying the choroidal melanomas is brightly hyperautofluorescent. Adjacent dark hypoautofluorescent areas surround the bright lipofuscin hyperautofluorescence. These may represent shifting and clumping of RPE cells or RPE atrophy (Lavinsky et al. 2007; Singh et al. 2010; Shields et al. 2007). Detection of lipofuscin by FAF imaging could play a role in the early detection of choroidal melanoma. Notably, orange pigment is considered one of five factors predictive of growth of small choroidal melanocytic tumours into melanoma. Subretinal fluid represents a physical barrier between the outer segments of photoreceptors and the RPE, thereby preventing normal phagocytosis of the shed outer segments of the photoreceptors. The latter subsequently accumulate on the outer retinal surface and in the subretinal space and are a source of autofluorescence. Histologically, the debris in the subretinal fluid consists of rounded or elongated densely packed multilamellar bodies that appear to result from the breakdown of outer segment discs. The retina anterior to the subretinal material shows attenuation of the photoreceptor layer. The subretinal material is bound posteriorly by the RPE (Singh et al. 2010). Shields et al. examined choroidal melanomas characterized by overlying orange pigment and subretinal fluid and found that the pigment corresponded to lipofuscin in macrophages located in the subretinal space. They also demonstrated that the fluid rim was slightly more hyperautofluorescence than the fluid centre (Shields et al. 2007; Singh et al. 2010). Following treatment (plaque radiotherapy combined with transpupillary thermotherapy), choroidal melanomas may show increased FAF due to an increase in lipofuscin and hyperpigmentation (Gündüz et al. 2007). All treated melanomas enrolled in our study showed an increase in FAF irrespective of the type of treatment (plaque radiotherapy, transpupillary thermotherapy or a combination of them) (Fig. 1). In conclusion, choroidal melanomas may be surrounded by confluent hyperautofluorescence. Most nevi are not hyperautofluorescence (Gündüz et al. 2009). Therefore, FAF studies may be a useful non-invasive tool with which to identify lipofuscin in pigmented choroidal lesions, especially nevi and small choroidal melanomas.