Differences among various standards with regard to slow crack growth measurement in multispecimen R- curve studies are outlined. An estimation of error in CTOD toughness due to inaccuracy in crack size measurement is given. An analytical approach has been used to optimize the number of measurement for typical curved crack fronts. In the case of relatively smooth crack front, nine point measurements as recommended by ASTM standard are found to be adequate. However, more measurements are necessary when the materials inhomogeneity and tunnelling effects contribute to irregular crack fronts. This has been evidenced both from direct measurement and analytical approach.