Abstract It is argued that the ‘routines as genes’ analogy is misleading in several respects. Neither genes nor routines program behaviour, if this is taken to involve, first, that they determine behaviour and, second, that they do so in a way that excludes conscious, deliberate choice. On a proper understanding of ‘gene’ and ‘routine’, knowledge of genes and routines falls far short of predicting behaviour. Furthermore, conscious, deliberate choice is not ruled out when genes or routines are operating. There is a sense in which it can be maintained that genes are (or act as) programs and that individual behaviour is based on them. Such programs might display considerable stability, but their causal impact on behaviour is so remote and indirect that knowing them has little predictive power. It might be possible to identify programs also at levels of organization higher than that of genes that have greater predictive power, but such programs are likely to be unstable over time. On a non-inflationary understanding of ‘routines’, individual organization members can be viewed as programs on which the smooth functioning of routines is based. This is a far cry from the claim that routines determine firm behaviour, let alone from the claim that they are key success variables in explaining how well (in terms of profitability) firms perform.