This study was designed to determine (1) the ovarian follicular dynamics in 12 Yorkshire X Landrace crossbred gilts by using as measures the type of ovaries (grape 'G', honeycomb 'H' and intermediate 'I' types) and their amplitude of change at different periods before puberty; (2) the relationship between ovarian types and serum concentrations of sex steroid (oestradiol-17b, E(2); testosterone, T; androstenedione, A) and luteinizing hormone (LH). Five successive laparoscopies at 5-day intervals between 160 and 180 days of age were performed on each gilt to determine the ovarian type and the distribution of follicles present on the ovarian surface. Blood sampling for RIA analyses of hormones was at 15-min intervals for 6 h in eight of the 12 gilts, 1 or 2 h before each laparoscopy. Within a gilt, no ovary maintained the same ovarian type over the five successive laparoscopies. The H type was found to be stable for three successive laparoscopies, whereas the G and I types were generally maintained for only two successive laparoscopies. When an ovary switched from the G type to the H or I type, the number of large follicles (greater than or equal to 6 mm in diameter) decreased (P<0.01) by seven follicles or less while that of the small follicles (1-3 mm in diameter) increased (P<0.01) by 15 follicles or more. When an ovary switched from the H type to the G or I type, the inverse phenomenon was observed; the number of large follicles increased (P<0.01) by six follicles or less while that of the small follicles decreased (P<0.01) by 16 follicles or more. Moreover, the concentrations of E(2), T, A and LH in serum were not influenced by ovarian types. However, when H gilts became I gilts, the concentration of A increased by 1 ng ml(-1). When I gilts became H gilts, the inverse phenomenon occurred; the concentration of A decreased by 0.68 ng ml(-1). The results of the present study demonstrate that between 160 and 180 days of age (1) all ovaries undergo morphological change at least once; (2) the time during which the three types of ovaries remain stable rarely exceeds 10 days for the H type, and 5 days for the G and I types; (3) the change of ovarian morphology is associated with a change in follicular population; (4) no relationship exists between ovarian types and serum E(2), T, A and LH concentrations. However, the concentration of A increases or decreases depending on whether H gilts change to I gilts or I gilts change to H gilts. Thus, during the prepubertal period in gilts, ovarian morphology is not static but undergoes continuous changes suggestive of follicular waves.