A standard procedure for measurement of fecal pH in dairy cows does not currently exist. Consequently, sample preparation may influence the precision of this measurement; thus, limiting comparisons across literature reports. The objectives of this study were to determine if differences exist based on preparation method, and to determine variation across methods. Thirty fresh fecal samples were collected from lactating Holstein cows housed in the same pen and consuming the same diet. Five samples were collected at a time and prepared according to the following methods: (1) direct measurement (DIR) in which the pH probe was directly inserted into the fecal sample; (2) strained fecal fluid (STR) obtained by squeezing the fecal sample through 4 layers of cheesecloth. Three dilution rates (distilled water:feces) were also tested: (3) 0.5:1 dilution (D1), (4) 1:1 dilution (D2), and (5) 2:1 dilution (D3). Each sample was prepared using all methods, resulting in a total of 150 pH measurements. The UNIVARIATE and GLM procedures of SAS were used to test normality and homogeneity of variance, respectively. The Shapiro-Wilk test confirmed that data were normally distributed (P = 0.08). The Levene’s test showed heterogeneity of variance (P = 0.02), thus the SATTERTHWAITE approximation of degrees of freedom for denominator was used for the ANOVA via the GLIMMIX procedure. Sample preparation method affected (P < 0.01) pH values, resulting in D3 having the highest pH of 6.91 ± 0.04, followed by D2 with a value of 6.79 ± 0.04. Measurements of pH by D1 and DIR were similar, and averaged 6.67 ± 0.04 (P = 0.17); whereas, STR had the lowest value of 6.60 ± 0.04. Descriptive statistics showed the standard deviation for the STR method was 0.173 and 0.174 for D2, while that of D1, D3 and DIR was 0.224, 0.226 and 0.296, respectively. These results demonstrate that pH measurements in strained fecal fluid or a 1:1 dilution rate have reduced variability when compared with direct measurements and other dilution rates.