The development of software systems for the biosciences is always clos
ely connected to experimental practice. Programs must be able to handl
e the inherent complexity and heterogeneous structure of biological sy
stems in combination with the measuring equipment. Moreover, a high de
gree of flexibility is required to treat rapidly changing experimental
conditions. Object-oriented methodology seems to be well suited for t
his purpose. It enables an evolutionary approach to software developme
nt that still maintains a high degree of modularity. This paper presen
ts experience with object-oriented technology gathered during several
years of programming in the fields of bioprocess development and metab
olic engineering. It concentrates on the aspects of experimental suppo
rt, data analysis, interaction and visualization. Several examples are
presented and discussed in the general context of the experimental cy
cle of knowledge acquisition, thus pointing out the benefits and probl
ems of object-oriented technology in the specific application field of
the biosciences. Finally, some strategies for future developments are
described.