Scientists often construct simplified and idealized models in order to study complex phenomena. Yet they do not model a phenomenon in its entirety but target only the aspects of the phenomenon which they consider relevant. Hence, the model is said to describe the target system and not the whole phenomenon. The term target system' has become popular in the philosophy of science, yet most authors do not provide a definition or analysis of the concept. The result is that the term is used ambiguously, which has undermined its potential value and usefulness for scientific practice. The aim of this dissertation is to provide a cogent account of target systems and their importance in science, with examples taken from case studies in ecology. The central issue I explore in my dissertation concerns the nature of target systems. What are target systems? How are they specified? How can they be evaluated? In my dissertation I give an account of target systems as real parts of systems in the world, which are specified through a process of partitioning and abstraction. I also provide a tentative theory of target system evaluation based on the notion of aptness for a particular scientific purpose. A deep understanding of nature and function of targets can resolve problems in science. I use the term target system analysis', to denote the specification of target systems of one enquiry and the comparison of targets across enquiries. The last part of the dissertation is devoted to the application of the theory of target system specification and evaluation to a case study from actual scientific practice, invasion biology. Target system reveals that a scientist constructing a unificatory framework in invasion biology faces a tradeoff between generality and predictability. A truly unified framework must incorporate a multitude of different causes of invasion, yet the causes of each invasion are unique. Hence, invasion biology can have a unified theory, based on the process of invasion, yet this theory will be of little use to predicting particular invasions.