When the definitive test strategy has been established in keeping with the development process estimated effort and planning, the development tests should be fleshed out further. This takes place through the allocation of test techniques to the various (groups of) units. The choice of test technique to be used depends on the selected depth of testing. Suitable test design techniques for development testing are, for example, the elementary comparison test and the decision table test. Techniques such as error guessing or checklists are equally applicable (see A Basic Set Of Test Design Techniques).

The integration of a system normally takes place in a number of steps, in which one or more units are integrated with each other and tested each time.

For the unit integration test, there are various possibilities as regards the sequence of unit integration:

• Top down - Testing takes place from the top downwards (e.g. working from within the menu screen). Units are replaced by stubs (see Types Of Test Tools).
• Bottom up - Testing is done from the bottom upwards (e.g. first data manipulation and only later the relevant screen). Units are replaced by drivers (see Types Of Test Tools).
• Available parts first - Integration purely based on the sequence of delivery of the units.
• Function by function - Integration based on the functions, as established in the functional design.

What the best integration sequence is and how many integration steps are necessary depends on the position of the most risk-related parts of the system. Techniques based on paths or equivalence classes are particularly to be recommended with integration testing, depending on the size of the parts to be integrated.