Our primary objective in this study was to determine if hemispherical (fish-eye) photographs could be used to estimate photosynthetic photon flux density (PPFD) in mature and old-growthPseudotsuga–Tsugaforests in the Oregon and Washington Cascade Range. LI-COR quantum sensors sampled PPFD at 10-s intervals at six points in each of two old-growth stands from January to December of 1991. Direct measures of PPFD were also made with photodiodes, which were calibrated against quantum sensors. They were used to sample PPFD for 6- to 14-day periods during the summer at 15 points in each of four stands, two mature and two old growth. Hemispherical photographs were taken at these sample points and were digitized and analyzed using the programCANOPY. The software predicts overall site openness and openness along the sun path. Regression models were developed based upon those parameters. The models predicted the mean daily PPFD for each month of the year. Two major factors were found that influenced regression models: stand age–structure (mature vs. old growth) and sky conditions (dry vs. wet seasons). Canopies in mature stands were more open than old growth stands, and their PPFD models were more heavily influenced by overall site openness. Conversely, PPFD models for old-growth canopies were more heavily influenced by openness along the sun path. Wet-season models were dominated by diffuse PPFD (overall site openness), and dry-season models were dominated by direct flux (openness along the sun path). Combining data sets from stands of comparable age–structure and seasons considerably increased regressionR2-values.