Though the realist account for ontological engineering has by now been described on text book level (Jansen/Smith 2008; Munn/Smith 2008), there is still considerable controversy about the impact and legitimacy of its philosophical background (Merrill 2010a, 2010b; Maojo et al. 2011), and what the methodology consists of in the first place. In the present paper, we try to summarize the realist paradigm in ten easy rules or commandments. We do not aim to present any new prescriptions; we rather try to compile and condense ontology design rules that have hitherto been presented at scattered places. We will argue that the way how they fit into the realist account follows from the realist conception of what an ontology is. As a result it will follow that realism, as argued for in this paper, is not an ideology, but an engineering approach for designing good ontologies which have a higher probability to be sustainable, interoperable, and adequate to their domain. Since the term ‘ontology’came into use in information science, it has been used for a variety of information artefacts. Both intension and extension of the term is debated. By different scientists, ontologies are said to be:(i) information artefacts,(ii) representations (eg, of conceptualizations),(iii) formal structures,(iv) theories,(v) hierarchies of types or universals. If we take a closer look at these competing characterisations, they turn out to be compatible, as they describe different aspects of ontologies. Ontologies can, eg, be representations and information artefacts at the same time, as the latter are nothing but artificial representations. Those who describe ontologies as formal structures, state the way how ontologies represent what they represent. Those who describe ontologies as theories, refer to what is the benchmark of what we represent, ie some theory about a certain domain (and many times it will be the best theory for that domain available to us). Those who, finally, describe an ontology as a hierarchy of types, universals, or classes, state what is represented in an ontology and what it is the benchmarking theories are theories of. As these characterizations of ontologies are complementary, we can put them all together. Hence, an ontology is an artificial representation (an information artefact, that is), that represents types or universals of a certain domain and the relations that hold according to a certain theory in a formal structure. More specifically, scientific ontologies are information artefacts that represent types of things and their relations of a certain domain according to the best available scientific theory, in order to support knowledge storage, processing and eliciting in the sciences. An immediate consequence of this definition is that an information artefact that is not intended to be useful for science cannot be considered to be a scientific ontology. We will now present our set of ten commandments. As in the biblical paradigm, we start with three ‘stage-setting’commandments that characterize our general approach, while the remaining seven commandments will lay out the details for the design of classes and relations within an ontology.