This study compared fullframe fisheye photography and cover photography with destructive leaf area index (L) estimation and the Licor LAI-2000 plant canopy analyser (PCA) in plantations of the vertical leaved species Eucalyptus globulus. Fullframe fisheye photography differs from circular fisheye photography in that the images have reduced field of view such that the zenithal range of 0-90 degrees extends to the corners of the rectangular image, roughly doubling image resolution compared to circular images. Cover images instead are obtained by pointing a 70 mm equivalent focal length lens (in 35 mm format) straight upwards. Measurements of cover and indirect estimates of plant area index (L,) were made in 12 stands of 6-8 years old Eucalyptus globulus. L was measured using destructive sampling and allometry in nine of these stands and ranged from 2.5 to 6.6. Both foliage cover and 4 from the PCA were well correlated with L from allometry, but fullframe fisheye photography provided poor estimates of L despite corrections for foliage clumping. Sampling location had a significant effect on estimates of crown porosity, crown cover and zenithal clumping index from cover photography. The zenithal extinction coefficient (k), calculated from L, crown porosity and cover, ranged from 0.14 to 0.25 and appeared to decrease as L increased; hence, we were unable to obtain an unambiguous estimate of k for E. globulus stands. Nonetheless, the study showed that L can be estimated from foliage cover with similar certainty to that of the PCA. We conclude that the greatest challenge facing indirect estimation of L in forests using photographic methods is to separate the effects of foliage angle from those of foliage clumping. (C) 2007 Elsevier B.V. All rights reserved.