Body size is often positively associated with mating success in fish. Howev
er, mating success may not accurately represent reproductive success (numbe
r of progeny produced). We tested the null hypothesis that reproductive suc
cess in male Atlantic cod, Gadus morhua, is not directly proportional to bo
dy size. Two allozyme loci and one minisatellite locus were used to estimat
e the proportion of larvae sired by pairs of males differing in body size a
nd spawning with one female in each of eight tanks during an entire spawnin
g season. Genotypes of progeny from 51 spawning events (egg batches) showed
that both males fertilised eggs in each tank. Multiple paternity of larvae
occurred in 35 batches. The estimated proportion of larvae sired by each m
ale varied randomly among egg batches within tanks. Male size, condition fa
ctor, and total or relative body-weight loss over the season were not corre
lated with the estimated proportion of larvae sired by each male during the
spawning season. Relative activity of males was marginally associated with
male reproductive success, with more active males siring a larger proporti
on of larvae. Male reproductive success was affected by female size, with m
ales much larger(> 25% total length) than females siring a smaller proporti
on of larvae.