Abstract It has been suggested that pine needles protect themselves against UV-B radiation via anatomical and chemical changes in the epidermal layer. This could lead to growth reductions if more assimilates are allocated to the protection mechanisms in the epidermis at the expense of the photosynthetic area. If the protection is insufficient, the UV-B radiation could cause oxidative stress. Moreover, the UV-B-induced stress could increase year by year and the responses could become cumulative. Therefore, two different UV-B field experiments were conducted at two sites. The first was a long-term experiment on mature, naturally growing Scots pines (Pinus sylvestris L.) in Oulu (65°N), Finland, during the growing seasons 1996–1998 (25% ozone depletion) and the second one was a one-growing-season experiment on Scots pine and loblolly pine (Pinus taeda L.) seedlings in Maryland (39°N), USA, in 1997 (16% and 25% ozone depletion scenarios). In Scots pine seedlings, UV-B treatments decreased the primary needle length at the early stages, but no significant differences were seen in fully-grown needles. In mature Scots pines, the response was opposite and UV-B accelerated needle growth in young needles. This suggests that at the early stage of development, the primary needles of seedlings are more sensitive than the fascicle needles of adult pines. Thus, the developmental stage of the studied organisms must be taken into account. Scots pine seedlings protected themselves against UV-B by increasing the thickness of the cutin layer and the epidermal wall (outer and periclinal walls and anticlinal wall), which features have been connected to xeromorphic structure, and by reducing significantly the cross-sectional, mesophyll and epidermal areas. In mature Scots pines, similar trends towards smaller cross-sectional and mesophyll areas and significantly reduced epidermal + hypodermal area were seen in fully-grown UV-B-treated needles. Loblolly pine responded to supplemental UV-B by producing more phenolic compounds into epidermal cell walls and epidermal lumen, which could have resulted in thicker outer epidermal walls. In both of the studied species, these protection mechanisms were sufficient to prevent growth reductions in fully-grown current-year needles. This demonstrates the difference between the studied species and shows that no general interpretations can be made on the effects of UV-B radiation on conifers. The cumulative effects of UV-B radiation were studied from current-year and current+2-year old needles of mature Scots pines during the third UV-B exposure season. The UV-B stress was observed in the current+2-year old needles of mature Scots pines as a degradation of total glutathione, GSHtot, and a bigger proportion of oxidized glutathione, GSSG%, in July during the third UV-B exposure season. In current-year needles, no increase in GSSG% was seen in either the first or the third season. After the third experimental season in September, the GSHtot levels decreased in UV-B-treated current-year needles. These results suggest that the effect of UV-B is duration-dependent and cumulative. The longevity of mature Scots pine needles at the study site is four to five years, and the effects of UV-B should thus be studied over the whole life span of a needle.