Scientific research on 'animal welfare' began because of ethical conce
rns over the quality of life of animals, and the public looks to anima
l welfare research for guidance regarding these concerns. The concepti
on of animal welfare used by scientists must relate closely to these e
thical concerns if the orientation of the research and the interpretat
ion of the findings is to address them successfully. At least three ov
erlapping ethical concerns are commonly expressed regarding the qualit
y of life of animals: (1) that animals should lead natural lives throu
gh the development and use of their natural adaptations and capabiliti
es, (2) that animals should feel well by being free from prolonged and
intense fear, pain, and other negative states, and by experiencing no
rmal pleasures, and (3) that animals should function well, in the sens
e of satisfactory health, growth and normal functioning of physiologic
al and behavioural systems. Various scientists have proposed restricte
d conceptions of animal welfare that relate to only one or other of th
ese three concerns. Some such conceptions are based on value positions
about what is truly important for the quality of life of animals or a
bout the nature of human responsibility for animals in their care. Oth
ers are operational claims: (1) that animal welfare research must focu
s on the functioning of animals because subjective experiences fall ou
tside the realm of scientific enquiry, or (2) that studying the functi
oning of animals is sufficient because subjective experiences and func
tioning are closely correlated. We argue that none of these positions
provides fully satisfactory guidance for animal welfare research. We s
uggest instead that ethical concerns about the quality of life of anim
als can be better captured by recognizing three classes of problems th
at may arise when the adaptations possessed by an animal do not fully
correspond to the challenges posed by its current environment. (1) If
animals possess adaptations that no longer serve a significant functio
n in the new environment, then unpleasant subjective experiences may a
rise, yet these may not be accompanied by significant disruption to bi
ological functioning. Thus, a bucket-fed calf may experience a strong,
frustrated desire to suck, even though it obtains adequate milk. (2)
If the environment poses challenges for which the animal has no corres
ponding adaptation, then functional problems may arise, yet these may
not be accompanied by significant effects on subjective feelings. Thus
, a pig breathing polluted air may develop lung damage without appeari
ng to notice or mind the problem. (3) Where animals have adaptations c
orresponding to the kinds of environmental challenges they face, probl
ems may still arise if the adaptations prove inadequate. For example,
an animal's thermoregulatory adaptations may be insufficient in a very
cold environment such that the animal both feels poorly and functions
poorly. We propose that all three types of problems are causes of eth
ical concern over the quality, of life of animals and that they togeth
er define the subject matter of animal welfare science.