Human immunodeficiency virus type-1 (HIV-1) relies on both partial and complete splicing of its full-length RNA transcripts to generate a distribution of essential spliced mRNA products. The complexity of the splicing process, which can employ multiple alternative splice sites, challenges our ability to understand how mutations in splice sites may influence the composition of the resulting mRNA pool and, more broadly, the development of viral progeny. Here, we begin to systematically address these issues by developing a mechanistic mathematical model for the splicing process. We identify as key parameters the probabilities that the cellular splice machinery selects specific splice acceptors, and we show how the splicing process depends on these probabilities. Further, by incorporating this splicing model into a detailed kinetic model for HIV-1 intracellular development we find that an increase in the fraction of either rev or tat mRNA in the HIV-1 mRNA pool is generally beneficial for HIV-1 growth. However, a splice site mutation that excessively increases the fraction of either mRNA can be detrimental due to the corresponding reduction in the other mRNA, suggesting that a balance of Rev and Tat is needed in order for HIV-1 to optimize its growth. Although our model is based on still very limited quantitative data on RNA splicing, Rev-mediated splicing regulation and nuclear export, and the effects of associated mutations, it serves as a starting point for better understanding how variations in essential post-transcriptional functions can impact the intracellular development of HIV-1. (c) 2005 Wiley Periodicals, Inc.