Abstract In recent years, several algorithms for Direct Digital Control (DDC) have been proposed in literature. Although some of these, such as PID or cascade controllers, are very commonly used in industrial applications, the more recent ones like optimal state feedback controllers using an observer or parameter adaptive controllers have rarely been applied to a real plant. The primary difficulty behind this application has been perhaps the lack of testing such algorithms on a pilot plant. Moreover, there has been no serious attempt to make a comparative study of the merits of such algorithms for an existing plant under actual operating conditions. In this paper, seven DDC algorithms are applied to the temperature control of a heat exchanger. These algorithms are: PID, cascade, compensation (pole assignment), deadbeat, half-proportional, adaptive and optimal state feedback controller using an observer. The system performance and sensitivity with respect to changes of the plant parameters, disturbances and set point variations are investigated for the heat exchanger using these algorithms. The results indicate that the more sophisticated algorithms, e.g. optimal state feedback, compensation and adaptive controllers, requiring more computer time and memory, yield relatively less improvement when applied to a low-order plant than do the simple algorithms such as PID or cascade. It was deduced that the PID controller with anti-windup is the most suitable one.