The abundance of old-growth associated herbs in young Pseudotsuga menziesii forests is limited in part by the low availability of propagules of these species, but it remains unclear whether the slow post-logging development of old-growth species in younger forests also reflects a lack of microhabitats suitable for seed germination. This study attempts to fill an almost total lack of knowledge about germination of understory species in second-growth Douglas-fir forests, and in young coniferous forests in general, while providing a preliminary exploration of the relative importance of these two limiting factors. The ability of six common forest herbs to germinate in a 40-year-old Pseudotsuga menziesii stand was tested, and survival and growth were monitored for seven years. Five of the six species germinated well (8–50% germination), while one open-site species failed to germinate. After seven years, survival of four old-growth associated species was ∼10% (of the original seed number), while only 1% of seeds of a disturbance-adapted species were still present as seedlings. For all species, seedlings grew slowly and did not reach reproductive size, forming a seedling bank analogous to that of shade-tolerant tree seedlings. All old-growth associated species showed higher overall seedling numbers or greater survival in gaps than in non-gap microsites, and two species showed higher survival in the more mesic lower part of the watershed. Naturally occurring seedlings were rare and included only easily dispersed species that are forest generalists or characteristic of mature, but not of old-growth forests. I conclude that germination of late-seral herb species in young Pseudotsuga menziesii forests is not strictly microsite limited, but that these species are instead limited by a combination of low propagule availability and low resource levels with concomitant slow growth rates.