Background Fish fin is a widely used, non-lethal sample material in studies using stable isotopes to assess the ecology of fishes. However, fish fin is composed of two distinct tissues (ray and membrane) which may have different stable isotope values and are not homogeneously distributed within a fin. As such, estimates of the stable isotope values of a fish may vary according to the section of fin sampled. Methods To assess the magnitude of this variation, we analysed carbon (δ13C), nitrogen (δ15C), hydrogen (δ2C) and oxygen (δ18O) stable isotopes of caudal fin fromjuvenile, riverine stages of Atlantic salmon (Salmo salar) and brown trout (Salmo trutta). Individual fins were sub-sectioned into tip, mid and base, of which a further subset were divided into ray and membrane. Findings Isotope variation between fin sections, evident in all four elements, was primarily related to differences between ray and membrane. Base sections were13C depleted relative to tip (~ 1 ‰) with equivalent variation evident between ray and membrane. A similar trend was evident in δ2H, though the degree of variation was far greater (~ 10‰). Base and ray sections were 18O enriched (~ 2‰) relative to tip and membrane, respectively. Ray and membrane sections displayed longitudinal variation in 15N mirroring that of composite fin (~ 1‰), indicating that variation in15N values was likely related to ontogenetic variation. Conclusions To account for the effects of intra-fin variability in stable isotope analyses we suggest that researchers sampling fish fin, in increasing priority, 1) also analyse muscle (or liver) tissue from a subsample of fish to calibrate their data, or 2) standardize sampling by selecting tissue only from the extreme tip of a fin, or 3) homogenize fins prior to analysis.