Neutron diffraction is a powerful tool for investigating interlayer coupling phenomena in magnetic superlattices. In the absence of interlayer spin coherence the diffraction spectrum has the form of a broad smooth maximum, i.e., it essentially reproduces the shape of the single-layer structure factor. In contrast, spectra from correlated superlattices exhibit characteristic fringe patterns. However, doubts are sometimes expressed as to whether the presence of fringes indeed provides meaningful evidence for interlayer coherence; such opinions are usually based on results of computer simulations that produce “fringelike” features even though the magnetization in consecutive layers changes in a random fashion. Clarity in this issue is certainly of importance, especially in studies of systems with antiferromagnetic layers, in which neutron diffraction is practically the only experimental tool capable of detecting interlayer spin coherence. We report the results of thorough computer modeling studies that show that the “pseudofringes” obtained from simulations on uncoupled superlattice models disappear after averaging the data from a sufficiently large number of random configurations. It is a reasonable expectation that in a real superlattice each magnetic layer breaks into a large number of domains, so neutrons passing through the sample “see” different situations. Hence, realistic modeling does require averaging over many random configurations, and our results point out that in following such a procedure, the “fringelike” features eventually wash out, and the simulated spectra become fully consistent with the analytical single-layer structure factor. However, there is still the possibility that in special situations the spectra from uncoupled systems may exhibit fringes. Suppose that the system symmetry allows discrete domain states (e.g., corresponding to in-plane easy axes), and that the average domain size and the superlattice period are both small in comparison to the coherence length of the incident neutron wave (since the coherence length in a typical experiment is of the order of 1000 A, this is not an unusual situation). Then, arguably, there is always a finite probability that equivalent domains located in different layers will fall into the “coherence volume,” and give rise to constructive interference, even though the domains are not coupled by any interactions. In view of that, our work included simulations on several different model systems that might, in principle, exhibit such an effect. Again, the averaged data from many thousands of runs showed no anomalies at the expected fringe positions, providing more evidence that neutron diffraction allows a fully reliable distinction of coupled and uncoupled superlattice systems.Neutron diffraction is a powerful tool for investigating interlayer coupling phenomena in magnetic superlattices. In the absence of interlayer spin coherence the diffraction spectrum has the form of a broad smooth maximum, i.e., it essentially reproduces the shape of the single-layer structure factor. In contrast, spectra from correlated superlattices exhibit characteristic fringe patterns. However, doubts are sometimes expressed as to whether the presence of fringes indeed provides meaningful evidence for interlayer coherence; such opinions are usually based on results of computer simulations that produce “fringelike” features even though the magnetization in consecutive layers changes in a random fashion. Clarity in this issue is certainly of importance, especially in studies of systems with antiferromagnetic layers, in which neutron diffraction is practically the only experimental tool capable of detecting interlayer spin coherence. We report the results of thorough computer modeling studies that show ...