Sir

The report on US laboratory animal populations1 contains the surprising assertion that there is an increasing use of animals in research in the United States and that lack of laboratory animal space is threatening the existence of valuable disease models. The proposed remedy includes an infusion of money to expand laboratory animal space.

Such US data as are available indicate that laboratory animal use is, in general, declining. It may be declining as dramatically as it has in Great Britain (more than a 50% decline since 1975), the Netherlands (50% since 1978), Switzerland (75% since 1983) or Germany (35% since 1989). Although the US data are incomplete because mice and rats are not tallied, despite making up 90% or so of the laboratory animals used, the annual report of dog, cat, primate, hamster, guinea pig and rabbit use shows that use of these six species has declined by about 40% since 1976. A report on laboratory animal use by the Department of Defense found that rat use declined by 49% and mouse use by 28% from 1983 to 1991 (ref. 2).

Another study pointed out that, among 20 commercial laboratories providing such data in their annual reports, rat and mouse use had declined by 39% from 1986 to 1994. Rat and mouse use at the National Institutes of Health (NIH) fluctuated between 1986 and 1993, but mouse use was highest in 1989 (at around 450,000) and then declined to 300,000 in 1991 before rising again to about 380,000 in 1993. Rat use declined slowly but steadily over the same period3. A 1966 report notes that in 1965 NIH used 670,000 mice and 163,000 rats4.

Although the available data also indicate that the use of mice (especially in the development of disease models using transgenic and knockout technology) has increased steadily over the past five to ten years, the claim that animal use in general is rising or that the lack of animal holding capacity has reached a crisis is not supported by either the data or by my personal experience of the space availability at four research institutions in the northeast United States from 1984 to 1997. Of course, if investigators now wish to set up facilities to maintain breeding colonies of new mouse models of human disease, then competition for animal space may well develop. Most American institutions are not set up to handle in situ breeding and colony maintenance because they usually purchase animals as needed from commercial suppliers.

The trends in biomedical research for the past 20 years have been away from animal use towards molecular and cellular investigation. The excitement about knockout and transgenic techniques may result in a reversal of the worldwide downward trend in laboratory animal use, but such a reversal is likely to be temporary. Like Sir Peter Medawar in 1969, we look to a future where our growing knowledge of biology — obtained through both animal and increasingly non-animal research — will “provide us with the knowledge that will make it possible for us, one day, to dispense with the use of [animals] altogether”5.

• The author's disagreement is with the findings of the US National Research Council, not with our report of those findings. — Editor, Nature.