The size-compositions of snapper, Pagrus auratus, caught using demersa
l trawls and longlines were compared for six sites sampled off Sydney,
N.S.W., Australia. We predicted that if gear-dependent selectivity oc
curred, it would be evident as changes to (1) the mean size of individ
uals, (2) the variance in size of individuals, or (3) both the mean an
d variance in size of individuals. Neither technique caught individual
s less than 90 mm fork length and catches exhibited temporal fluctuati
ons during the 4 years of sampling. Although trawls were more producti
ve and provided almost 65% of the entire catch, snapper caught on long
lines were significantly larger in size and covered a wider range of s
izes. These results were also manifest across sites despite the signif
icant spatial differences in the mean size of individuals, The differe
ntial gear selectivity also affected catches of undersized and legal-s
ized snapper, Trawls caught significantly smaller undersized and legal
-sized snapper compared with longlines, Several hypotheses were propos
ed to account for the absence of juvenile snapper in the catches of bo
th techniques, The results were consistent with our last prediction in
that both the mean and variance in size differed between trawls and l
onglines. Finally, the results are discussed in terms of their implica
tions for fishery-independent, standardized trawl surveys and the subs
equent use of the resultant data in stock assessment models.