Modifications in gene expression determine many of the phenotypic differentiations between closely related species. This is particularly evident in reproductive tissues, where evolution of genes is more rapid, facilitating the appearance of distinct reproductive characteristics which may lead to species isolation and phenotypic variation. Large-scale, comparative analyses of transcript expression levels have been limited until recently by lack of inter-species data mining solutions. Here, by combining expression normalisation across lineages, multivariate statistical analysis, evolutionary rate, and protein-protein interaction analysis, we investigate ortholog transcripts in the male accessory glands and testes across five closely related species in the Anopheles gambiae complex. We first demonstrate that the differentiation by transcript expression is consistent with the known Anopheles phylogeny. Then, through clustering, we discover groups of transcripts with tissue-dependent expression patterns conserved across lineages, or lineage-dependent patterns conserved across tissues. The strongest associations with reproductive function, transcriptional regulatory networks, protein-protein subnetworks, and evolutionary rate are found for the groups of transcripts featuring large expression differences in lineage or tissue-conserved patterns.