Leaf area index (LAI) was estimated at 15 sites in the Swiss Long-Term Forest Ecosystem Research Programme (LWF) in 2004-2005 using two indirect techniques: the LAI-2000 plant canopy analyzer (Licor Inc.) and digital hemispherical photography, applying several exposure settings. Hemispherical photographs of the canopy were analysed using Hemisfer, a software package that offers several new features, which were tested here: (1) automatic thresholding taking the gamma value of the picture into account; (2) implementation of several equations to solve the gap-fraction inversion model from which LAI estimates are derived; (3) correction for ground slope effects, and (4) correction for clumped canopies. In seven broadleaved stands in our sample set, LAI was also estimated semi-directly from litterfall. The various equations used to solve the gap-fraction inversion model generated significantly different estimates for the LAI-2000 measurements. In contrast, the same equations applied in Hemisfer did not produce significantly different estimates. The best relationship between the LAI-2000 and the Hemisfer estimates was obtained when the hemispherical photographs were overexposed by one to two stops compared with the exposure setting derived from the reading of a spotmeter in a canopy gap. There was no clear general relationship between the litterfall and the LAI-2000 or the hemispherical photographs estimates. This was probably due to the heterogeneity of the canopy, or to biased litterfall collection at sites on steep slopes or sites subject to strong winds. This study introduces new arguments into the comparison of the advantages and drawbacks of the LAI-2000 and hemispherical photography in terms of applicability and accuracy.