Due to a multiplicity of methods used to determine breeding systems in Angiosperms and the low comparative value of results derived from them, plant reproduction biologists have had limited capacity to integrate and discuss the available data properly and postulate general evolutionary trends of breeding systems in flowering plants. We propose a new analytical approach to classify breeding systems in Angiosperms combining statistical and conceptual criteria, which will allow us to describe and compare the available mating system data in a unified form, considering a broader spectrum of potentially possible reproductive strategies that have evolved or could evolve across the entire flora. The published data on breeding systems in flowering plants were re-evaluated using a novel procedure. Four breeding system indexes (BSI), with values ranging from zero to infinite, were analyzed using confidence intervals of relative risk ratio (RR), followed by a t-test to discriminate experimental values of RR from zero and 1.0. After that, for each index, data were assigned to one of five possible discrete categories. In addition to this, a measure of strength (M) of each category was determined. Indexes of inbreeding and outbreeding depression were also examined using the same statistical tests. In addition to this, pollination systems were assessed according to the outcomes of four experimental pollination tests evaluated using a novel framework. Published data from 1,908 taxa were used to re-analyze the breeding systems of Angiosperms. Non-agamospermous species dominated. The frequency distributions of categories associated with Spontaneous Self-pollination and Self-fertility indexes tended to be approximately unimodal, with a clear dominance of non-spontaneous self-pollinating and xenogamous species. In relation to the Self-incompatibility Index, the largest group of species evaluated was partially self-incompatible and exhibited inbreeding depression. The actual number of composite breeding systems inferred based on our analyses was only a small fraction (8.9%) of all the mathematically possible ones; however, a significant number of new reproductive strategies, not previously reported, were identified. Similarly, the observed combinations between results obtained from the four pollination tests and estimates of natural reproductive efficiency were only 22.1% of all the mathematically possible combinations. Within Angiosperms, there is a marked trend towards evolution of partial self-incompatibility with significant inbreeding depression. Potential and identified breeding systems are highly diverse according to our classification system. The combination of categories of four breeding system indexes, and the possibility to evaluate their strength, allows establishing the relative importance of each reproductive attribute for each examined species. This detailed dissection of breeding strategies will expand our knowledge on the complex evolutionary routes and patterns followed within the different clades of Angiosperms.