60 MHz undecoupled α-, 90 MHz undecoupled β-, and 90 MHz decoupled α- and β-proton spectra were measured for seven poly(vinyl chloride) samples, four of these being commercial type polymers, the others were prepared at +75°C, +5°C and −30°C. The results were used to assess the preferred approach to the determination of tacticity. The reproducibility of the values from different solutions of a given polymer is significantly worse than that between a series of runs on a particular solution, showing that sampling errors are larger than those from instrumental sources. Although the 60 MHz spectra were found to be unsatisfactory when interpreted on a first order basis, curve fitting methods, which were similar to those used for the other types of spectra and involved the use of reasonable assumptions for chemical shifts and coupling constants, gave results of comparable precision to those from the two types of decoupled spectra and superior to those from the undecoupled 90 MHz spectra. The preferred approach is to combine the results from α- and β-proton spectra and to examine more than one solution of a given material. The applicability of Bernoullian propagation statistics was assumed in the majority of the curve fitting calculations; in the cases where this was not so the results were used to test this assumption and it has proved to be valid.