Seven current contending cancer theories consider different sets of critical events as sufficient for tumorigenesis. These theories, most recently the microRNA dysregulation (MRD) theory, have overlapping attributes and extensive empirical support, but also some discrepancies, and some do not address both benign and malignant tumorigenesis. By definition, the most efficient tumorigenic pathways will dominate under conditions that selectively activate those pathways. The MRD theory provides a mechanistic basis to combine elements of the current theories into a new hypothesis that: (i) tumors arise most efficiently under stress that induces and sustains either protective or regenerative states of adaptive hyperplasia (AH) that normally are epigenetically maintained unless terminated; and (ii) if dysregulated by a somatic mutation that prevents normal termination, these two AH states can generate benign and malignant tumors, respectively. This hypothesis, but not multistage cancer theory, predicts that key participating AH-stem-cell populations expand markedly when triggered by stress, particularly chronic metabolic or oxidative stress, mechanical irritation, toxic exposure, wounding, inflammation, and/or infection. This hypothesis predicts that microRNA expression patterns in benign vs. malignant tumor tissue will correlate best with those governing protective vs. regenerative AH in that tissue, and that tumors arise most efficiently inmutagen-exposed stem cells that either happen to be in, or incidentally later become recruited into, an AH state.