Broad-leaved monocot herbs form one of the most common and diverse growth forms of Neotropical plants. Their significance and frequency of occurrence is particularly notable in the understories of tropical rainforests, where they form a dominant element. We assessed and quantified changes in the cover and diversity of understory herb communities in a chronosequence of 1 ha permanent plots established as part of a multidisciplinary study on tropical forest regeneration in the Atlantic lowlands of northeastern Costa Rica. Sampled were two young stands cleared 12 years ago, two secondary forests with 21 and 39 of years of recovery since clearance, and two stands in old-growth primary forest. Changes in species composition during succession were assessed using Chao’s Jaccard similarity index. Observed species richness ranged from 15 to 26 species in individual plots, with the greatest number of species in the 21-year intermediate-age and fewest in the young 12-year plots. Herb species sampled represented 6 families, 15 genera, and 39 species, with the Araceae contributing the largest number of species. Ten species were sampled in all six stands, while fourteen species were found exclusively in one plot. Herb density (ramets m−2) showed a hump-shade trend, with peak density in the intermediate stands and a lower level in mature and young secondary forests. Mean herb cover in 25 m2 quadrats ranged from 2.0% (young stand) to 22.7% (intermediate-age stand) and differed significantly both among stand types and among sites. Both observed and estimated species richness increased along the chronosequence as a whole, with the highest number of species in primary forest, although only slightly higher than in intermediate-age stands. Over half of the species exhibited some degree of clonal growth, with the extent of clonal spread varying among species and forest stands. Although we did not find a clear pattern between clonality and forest age, we observed a greater number of clonal patches in secondary over primary forest stands.