A proton NMR method, based on simple Bloch-decay spectra in the solid state, is presented that enables one to follow, with excellent sensitivity, structural changes associated with aging in semicrystalline polymers whose T, is well below the aging temperature. The method is demonstrated for two representative isotactic polypropylene (iPP) samples, a Ziegler-Natta product and a metallocene product. Starting with samples that had been melt crystallized at a cooling rate of 1 degreesC/min and then aged at ambient temperature for long periods of time, subsequent mild heating cycles between ambient temperature and temperatures below 90 degreesC were applied. Such heating cycles remained more than 70 degreesC below the major crystalline melting temperature for iPP. Aging at 20 degreesC was monitored by NMR over aging times, 6 min < t(age) < 4 d, following those heating cycles. It was shown that changes in the Bloch-decay spectra, corresponding to a lower limit of 2-3% of the mass of iPP being transformed from mobile to rigid components, accompanied the aging process over the 4 d period. Moreover, the time dependence was linear in 109(tag,). It was further shown that the population of those noncrystalline (NC) stems with the highest mobility was most strongly reduced by aging; this observation does not, however, unambiguously establish that these same chains were the actual stems participating in the newly formed structures. The aging process was also shown to be reversible in the sense that the structures formed could be completely destroyed by repeating the mild heating cycle. Attention was paid to the definition of crystallinity, and an operational definition of crystallinity for the NMR measurements was based on the component with a long (> 150 ms) value of T-1xy, the relaxation along the quantization axis of the toggling frame in a multiple pulse (MP) experiment. The complementary NC component, by this definition, includes not only the motionally averaged protons seen in the Bloch-decay spectrum but also certain protons with more hindered motions which, by Bloch-decay criteria, appear rigid. Approximately 75% of the NC protons transformed by aging are converted to "crystalline" protons, using the T-1xy definition; hence, the structural changes in aging seem to be dominated by crystallization, which we are comfortable to call secondary crystallization. A quick assay of the longitudinal proton relaxation, T-1(H), was also made during aging. Aside from some changes originating from oxygen losses during heating, T-1(H) was found to be independent of aging time, and implications are pursued. Without making firm conclusions, observations are noted that may have relevance to the morphological location of secondary crystallites and to the factors that influence the amount of material available for such crystallization. The extent of secondary crystallization in the metallocene iPP is only modestly smaller than in the Ziegler-Natta iPP, although the metallocene iPP has a lower stereo-regiodefect concentration, a narrower polydispersity, and no expected "amorphous fraction". It is speculated that significant secondary crystallization would also characterize a defect-free iPP with low polydispersity.