(1) Cirsium palustre, Galeobdolon luteum and Sanicula europaea were used in factorial experiments to investigate the effects of two features of vegetational shade: shade density, measured as the percentage transmission of photosynthetically active radiation (PAR), and shade type, measured as the ratio of red to far-red radiation (R/FR ratio). Neutral shade has an unaltered R/FR ratio, filtered shade a low R/FR ratio as under a deciduous tree canopy in summer. (2) The growth of all species was reduced in shade but not proportionately to shade density, partly because of compensatory increases in specific leaf area. Shade type had little or no effect. Plants died under 1 % PAR: Cirsium quickly, Galeobdolon more slowly and Sanicula most slowly of all, but in all cases more rapidly in filtered than in neutral shade. All grew well at 20% PAR and above. (3) Cirsium leaves and Sanicula petioles were longer the less dense the shade and tended to be longer in filtered shade. The lengths of Galeobdolon internodes and petioles were greater in denser shade and in filtered shade. (4) The flowering of Galeobdolon was reduced in filtered shade as well as in dense shade and independently of plant size. The flowering of Sanicula was not affected by shade type but was related to shade density and plant size. (5) Herbaceous woodland plants respond to filtered shade although increased height growth to overtop the shading plants would not be appropriate. Leaf display responses and increases in leaf and petiole lengths help to optimize PAR interception among the ground vegetation.