This study was conducted to help assess whether recent harvest rates o
f ducks banded before the hunting season in the United States varied a
mong stocks (i.e., species, age-sex cohort, location of banding) and w
hether regulatory actions have been successful at shifting harvest pre
ssure among stocks in the desired manner. Direct recovery rates during
1976-91 were used to index harvest rates and were influenced to some
degree by location of banding, age class, sex, species, and hunting re
gulations. However, differences in recovery probability among species
were consistent (P = 0.787) over 3 periods with different hunting regu
lations (1976-84, 1985-87, 1988-91), despite regulatory changes design
ed to affect species differentially. Band-recovery models that exclude
d the interaction between hunting regulations and sex of ducks were in
adequate (P less than or equal to 0.013), suggesting sex-specific chan
ges in recovery probabilites with changes in regulations. There was li
ttle evidence (P less than or equal to 0.151) that regulatory changes
affected recovery rates of ducks from 3 banding reference areas differ
entially. Harvesting duck stocks in an optimal manner requires an abil
ity to harvest selectively, a good understanding of each stock's dynam
ics, and knowledge of any interdependence in stock sizes.