Crossbred progeny from females, 3 yr old and older, representing Brown Swiss (B), Red Poll (R), Hereford (H) and Angus (A) maternal grandsires and maternal granddams in a diallel crossing experiment were evaluated to estimate breed mean maternal heterosis (mean hMi) for each breed, maternal heterosis for specific breed cross females (hMij), average maternal heterosis for all crosses (mean hM) and breed grandmaternal effects (gM'i) for preweaning calf traits. Estimates of mean hM were significant for birth date and weight, 200-d weight and 200-d weight/cow exposed. Even though calves with crossbred dams weighed more at birth, they did not differ from calves with straightbred dams in frequency of calving assistance. The estimates of hMij were similar for most crosses and most traits, except 200-d weight and 200-d weight/cow exposed. The largest heterotic advantage was exhibited by progeny of BH reciprocal cross females, which exceeded crossbred progeny from BB and HH females for 200-d weight/cow exposed by 35.9 kg. The smallest heterotic effect was exhibited by crossbred progeny of RA reciprocal cross females. Progeny with R maternal granddams exhibited a higher (P less than .05) live calf born and weaned percentage than progeny with H maternal granddams. Progeny with B maternal granddams were born later (P less than .05) in the calving season than progeny with R maternal granddams. Progeny with A maternal granddams exhibited a lower frequency of calving assistance than progeny with H maternal granddams, and progeny with H maternal granddams were heavier (P less than .05) at 200 d than progeny with R maternal granddams. There was a tendency for estimates of breed grandmaternal effects (gMi) to exhibit an inverse relationship with estimates of breed maternal effects (gMi).