This analysis uses information obtained primarily from museum collecti
ons to estimate demographic parameters of populations of thrushes (Tur
dus spp.) from throughout the Western Hemisphere. Adult survival rates
were estimated from the proportions of individuals >1 yr old in museu
m collections, assuming constant population size, discrete reproductiv
e seasons, unbiased collecting, and maturity at 1 yr old. The effects
of relaxing each of these assumptions were examined. The resulting bia
ses were found to be small compared to differences between populations
, and in several cases, they were opposite to the conventional wisdom
that survival rate increases towards the equator. The annual survival
rate of adults (S) is estimated without bias by E(S) = A/(A + I), whic
h has a standard error of [(AI)/(A + I)(3)](1/2), where A is the numbe
r of adults in a collection and I is the number of immatures. In Turdu
s, adult and immature birds can be distinguished by the appearance of
the secondary coverts, the immature forms of which are retained until
the first postnuptial molt at approximate to 1 yr old. The present ana
lysis included 8653 specimens from 30 populations of 19 species from A
laska to southern Patagonia. Additional data tabulated from museum spe
cimens included numbers of juveniles, evidence of breeding, worn pluma
ge, and molt. Estimated annual survival rates varied from a mean of 0.
56 in temperate North American populations, to 0.68 in subtropical Sou
th American populations, 0.76 in lowland tropical populations, and 0.8
0 and 0.85 in tropical montane populations in Central and South Amer i
ca, respectively. Samples of adults and immatures restricted to the pe
riod immediately prior to the breeding season gave results that were i
ndistinguishable from samples summed over the entire year. Survival ra
te was most strongly correlated with the difference between maximum an
d minimum mean monthly temperatures (r = -0.84). Thus, survival is inv
ersely related to seasonality of temperature. An index to fecundity (J
) was estimated by the frequency of juvenile specimens (generally <2 m
o old) relative to those of adults in museum collections. This index w
as inversely related to annual adult survival. A relative index of pre
reproductive (generally Ist-yr) survival (S-a) was calculated as the a
nnual adult mortality [1 - E(S)] divided by J. Estimated S,was indepen
dent of estimated S. Calculations of annual fecundity (F) from field s
tudies on breeding of selected species supported both of these observa
tions. These results, which are consistent with conclusions based upon
broader comparisons among birds, suggest several hypotheses: (1) adul
t survival in Turdus populations is determined by physical conditions
during the nonbreeding season, particularly various ecological effects
of low temperature; (2) fecundity is sensitive to density-dependent f
eedbacks from adult population size and, possibly, to environmental fa
ctors correlated with other factors that independently affect adult su
rvival; and (3) Ist-yr survival, age at maturity, or both, also are af
fected by density-dependent factors.