Native starches with wide varying amylose content were processed by injection molding. The injection-molded materials were conditioned in water for 20 days and sealed in glass capillaries. Simultaneous wide- and small-angle X-ray scattering (WAXS and SAXS, respectively) were recorded during thermal heating using a synchrotron source. Crystallinity, SAXS invariant, Q, and long period, L, were measured as a function of heating temperature. The injection-molding process provokes a destruction of the crystal forms A (cereal starch) and B (tubercle starch) but favors a development of the crystal form V, After wet conditioning, WAXS of the injection-molded samples shows again the appearance of the crystal forms A or 13, and crystallinity reaches values similar or larger than those of native starch. A constant heating rate (5degreesC/min) was particularly used for a comparison of potato and corn starch with a similar amylose content. While the crystallinity associated to forms A and B slowly decreases below 55degreesC and then rapidly decreases until its disappearance at 85-90degreesC, the invariant shows a maximum around 40degreesC and rapidly decreases thereafter. The total nanostructure disappearance occurs at temperatures about 10degreesC higher for the case of potato starch. In addition, a recovery of the WAXS and SAXS maxima during the subsequent cooling process before reaching room temperature was observed only for potato starch. Analysis of WAXS and SAXS for the rest of the starch materials reveals clear differences in the structural parameters of the samples that cannot be easily explained solely on the basis of the amylose content. Thus, for Cerestar and Roquette, it is noteworthy that there was a continuous decrease of L until its total disappearance as well as the persistence of crystallinity (form B), presumably stabilized by the presence of the V-h structure (12-15%). Real-time crystallization experiments on two amorphous injection molded samples, waxy maize (free amylose starch) and potato starch, are also discussed. It is shown that the absence of amylose delays the recrystallization of amylopectine during the experiment. (C) 2004 Wiley Periodicals, Inc.