The structure and performance of the locally optimum detectors for composite signals in multiplicative noise are investigates. A generalized model with which we can represent multiplicative noise as well as additive noise is considered for signal detection problems. The signals considered here are composite signals which contain both deterministic signal components and stochastic signal components. The locally optimum detector consists of a nonlinearity followed by an accumulator, whose output is compared with a threshold. The detection nonlinearties for specific probability density functions are obtained to show the detection structures. To illustrate that performance of the locally optimum detectors, finite sample-size performance characteristics are obtained and compared with those of other detectors.